# Data dictionary

Public evidence package v1.0.2. Every public file, what it is, the endpoint it serves, and every field. Units: proportions are 0–1; pp means percentage points; times are UTC ISO 8601 unless stated. Placeholders in field paths: `{series}`, `{cycle}`, `{engine}`, `{window}`, `{sha256}` (answer hash) and `{key}` (a code, agent or composite key named in the description). Allowed values are listed after a colon and separated by `|`.

Shared value sets:

- series: uae-glp1 (Wegovy, UAE; registered) | singapore-sglt2 (Jardiance, Singapore; registered) | saudi-t1d (Tzield, Saudi Arabia; post-registration, D10)
- engine: perplexity (Perplexity, sonar) | google (Gemini, gemini-2.5-flash) | openai (ChatGPT, gpt-4o-mini) | anthropic (Claude, claude-haiku-4-5)
- window: before-v4 | before-v5 | during | after (protocol §3); post-cutoff never appears in an admitted cycle
- cycle code: UAE-Cnn | SG-Cnn | SA-Cnn, a neutral identifier of one scan of one series' battery on four engines (the study's internal cycle identifiers are withheld, R5a; fields keep their frozen names, such as slug and cycleSlug)
- Answer hash: the sha256 of a stored answer capture. It identifies an answer without its text; the held pull can be matched to it.
- Opaque question identifier (R2e): `<uae|sg|sa>-<f|q><NN>`. `f` marks a focused question (excluded from the E2 naming population), `q` a brand-neutral question. Stable within a series across every file of this package; it says nothing about the question's wording.
- Withheld page token: `[<type>-NN]` for an entity and `[<type>-NN]/pNNN` for one of its pages; `[platform-page:<platform>]/pNNN` for a platform page.

## `addenda/PN1-naming-panel-11.csv`

- **Role:** addendum-PN1-naming-panel-11. Daily naming counts by engine (Perplexity, ChatGPT, Claude) for the monitoring panel without its weight-management category, computed after publication from the stored panel readings by the measure of the monitoring-panel rows of addenda/PN1-naming.csv; no question or answer text.
- **Endpoint:** POST-PUBLICATION NOTE PN1, corrected (v1.0.2): monitoring-panel naming in its 11 categories that track no diabetes or weight-management medicine, 20 September – 13 October 2026
- **Source:** `docs/research/easd-2026/addenda/PN1-naming-panel-11.csv` (committed, not frozen); byte-identical to the source

| Column | Meaning, units and allowed values |
|---|---|
| `date` | UTC date |
| `stream` | PN1 addenda: monitoring-panel-11 (the monitoring panel without its weight-management category: its 11 categories that track no diabetes or weight-management medicine; v1.0.2) |
| `engine` | PN1 addenda: perplexity (Perplexity, sonar) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5); Gemini is not included |
| `cycle_admitted` | PN1 addenda: empty (the monitoring panel has no study cycles) |
| `named` | PN1 addenda: dictionary entity slots named in repeat-0 answers (one answer per question per day), summed over the 11 categories |
| `scored` | PN1 addenda: dictionary entity slots checked, summed over the 11 categories |
| `share_named` | PN1 addenda: named / scored, proportion |

## `addenda/PN1-naming.csv`

- **Role:** addendum-PN1-naming. Daily aggregates by stream and engine (Perplexity, ChatGPT, Claude), computed after publication from stored answer captures and stored naming counts; no question or answer text.
- **Endpoint:** POST-PUBLICATION NOTE PN1 (v1.0.1): Perplexity response structure and naming, 20 September – 13 October 2026
- **Source:** `docs/research/easd-2026/addenda/PN1-naming.csv` (committed, not frozen); byte-identical to the source

| Column | Meaning, units and allowed values |
|---|---|
| `date` | UTC date |
| `stream` | PN1 addenda: uae-glp1 \| singapore-sglt2 \| saudi-t1d (the study series) \| monitoring-panel (a fixed daily panel of 60 questions across 12 product categories, outside the study; one category concerns weight-management medication and tracks diabetes medicines; the other 11 track no diabetes or weight-management medicine) |
| `engine` | PN1 addenda: perplexity (Perplexity, sonar) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5); Gemini is not included |
| `cycle_admitted` | PN1 addenda: yes \| no for a study cycle (supplement §4.6); empty for the monitoring panel |
| `named` | PN1 addenda: study series, scored answers naming the medicine (the registered E2 measure, as stored); monitoring panel, dictionary entity slots named in repeat-0 answers (one answer per question per day) |
| `scored` | PN1 addenda: study series, scored answers; monitoring panel, dictionary entity slots checked |
| `share_named` | PN1 addenda: named / scored, proportion |

## `addenda/PN1-response-structure.csv`

- **Role:** addendum-PN1-response-structure. Daily aggregates by stream and engine (Perplexity, ChatGPT, Claude), computed after publication from stored answer captures and stored naming counts; no question or answer text.
- **Endpoint:** POST-PUBLICATION NOTE PN1 (v1.0.1): Perplexity response structure and naming, 20 September – 13 October 2026
- **Source:** `docs/research/easd-2026/addenda/PN1-response-structure.csv` (committed, not frozen); byte-identical to the source

| Column | Meaning, units and allowed values |
|---|---|
| `date` | UTC date |
| `stream` | PN1 addenda: uae-glp1 \| singapore-sglt2 \| saudi-t1d (the study series) \| monitoring-panel (a fixed daily panel of 60 questions across 12 product categories, outside the study; one category concerns weight-management medication and tracks diabetes medicines; the other 11 track no diabetes or weight-management medicine) |
| `engine` | PN1 addenda: perplexity (Perplexity, sonar) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5); Gemini is not included |
| `cycle_admitted` | PN1 addenda: yes \| no for a study cycle (supplement §4.6); empty for the monitoring panel |
| `answers` | answers (count); in Figure 11, the row's answer group: citing the page \| not citing \| before \| after \| a decomposition or echo label; in addenda/PN1-response-structure.csv, the answers stored for that stream, engine and day |
| `mean_results_per_answer` | PN1 addenda: mean count of search results (Perplexity) or cited sources (ChatGPT, Claude) returned per answer; the objects differ by API, so engines are not comparable with one another |
| `share_exactly_10_results` | PN1 addenda: share of answers carrying exactly 10 results, proportion |
| `mean_distinct_domains_per_answer` | PN1 addenda: mean count of distinct registrable domains among an answer's cited links |
| `mean_characters` | PN1 addenda: mean answer length, characters |
| `mean_inline_markers` | PN1 addenda: mean count of in-text citation markers of the form [n] per answer |
| `share_with_headings` | PN1 addenda: share of answers containing at least one markdown section heading, proportion |

## `analysis/easd-2026-analysis.json`

- **Role:** analysis. The single registered analysis of 11 Oct 2026 with the D21 errata and D23 self-check, at full precision.
- **Endpoint:** E2 (registered, UAE; Saudi post-registration), E3, E4 (quarantined), E5, Saudi E3 and E3-i, bridge self-check
- **Source:** `docs/evidence/computed/easd-2026-analysis.json` (frozen); transformed (see REDACTION-LOG.md)

| Field | Type | Meaning, units and allowed values |
|---|---|---|
| `.deterministic` | boolean | true: the run writes no wall-clock field |
| `.generatedBy` | string | provenance string (script, protocol section) |
| `.protocol` | string | provenance string (script, protocol section) |
| `.interpretations[].id` | string | I1–I5 (D16) |
| `.interpretations[].text` | string | the reading, verbatim |
| `.inputs.pull.pulledAt` | string | UTC time of the held pull |
| `.inputs.pull.from` | string | first date pulled (UTC) |
| `.inputs.pull.to` | string | last date pulled (UTC) |
| `.inputs.coding` | string | path of the coding input |
| `.inputs.saudi` | string | path of the Saudi coding input |
| `.inputs.bridge` | string | path of the bridge input |
| `.inputs.e1` | string | path of the E1 input |
| `.scorerValidation.cyclesChecked` | number | cycles whose answers were re-scored by the production scorer and matched to the stored governed counts |
| `.scorerValidation.mismatches[]` | (empty) | cycles where the re-score differed (empty: none) |
| `.perCycleNaming[].series` | string | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `.perCycleNaming[].slug` | string | cycle code: UAE-Cnn \| SG-Cnn \| SA-Cnn, a neutral identifier of one scan of one series' battery on four engines (the study's internal cycle identifiers are withheld, R5a; fields keep their frozen names, such as slug and cycleSlug) |
| `.perCycleNaming[].scannedAt` | string | UTC completion time of the cycle |
| `.perCycleNaming[].date` | string | UTC date of the cycle |
| `.perCycleNaming[].window` | string | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `.perCycleNaming[].engineVersion` | number | instrument version: 4 or 5 (§2) |
| `.perCycleNaming[].byEngine.{engine}.k` | number | E2: governed answers naming the measured medicine (count) |
| `.perCycleNaming[].byEngine.{engine}.n` | number | E2: governed answers (count; focused questions excluded) |
| `.e2[].series` | string | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `.e2[].registered` | boolean | true for the registered UAE arm; false for the Saudi arm (D10) |
| `.e2[].engine` | string | engine: perplexity (Perplexity, sonar) \| google (Gemini, gemini-2.5-flash) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5) \| all (four engines pooled) |
| `.e2[].beforeRef.k` | number | naming count of the pre-congress v4 reference |
| `.e2[].beforeRef.n` | number | answers of the v4 reference |
| `.e2[].beforeRef.rate` | number | k/n, 4 decimals |
| `.e2[].beforeRef.cycles` | string | which v4 cycles define the reference (I2) |
| `.e2[].{window}.k` | number | pooled naming count in the window (during, after) |
| `.e2[].{window}.n` | number | pooled governed answers in the window |
| `.e2[].{window}.rate` | number or null | k/n, proportion, 4 decimals |
| `.e2[].{window}.clusterCI95[]` | number | question-clustered bootstrap 95% interval, proportion (robustness, D16) |
| `.e2[].{window}.questions` | number | question clusters in the bootstrap |
| `.e2[].offset.pairedQuestions` | number | questions paired in the bridge |
| `.e2[].offset.offset` | number | bridge offset v5 − v4, proportion |
| `.e2[].offset.ci95[]` | number | bootstrap 95% interval of the offset, proportion |
| `.e2[].band[]` | number | registered v4 single-cycle range [low, high], proportion (§4; I5 for Saudi) |
| `.e2[].bridgedBand[]` | number | band + offset interval bounds (I1), proportion |
| `.e2[].movementDeclared` | boolean | registered E2 verdict: after rate outside the bridged band |
| `.e2[].magnitudePp` | number | after rate − bridged pre-congress reference, percentage points (I2) |
| `.e2[].magnitudeLabel` | string | SMALL \| POTENTIALLY RELEVANT \| DECISION-RELEVANT \| MAJOR (house scale) |
| `.e2[].beyondBandPp` | number | distance beyond the nearer band edge, percentage points, 1 decimal; 0 inside |
| `.e2[].precision.afterRateExact` | number | after k/n unrounded (D19) |
| `.e2[].precision.bridgedAsDecided[]` | number | bridged band as used for the verdict |
| `.e2[].precision.distanceAsDecidedPp` | number | signed distance to the nearer edge, pp; negative inside (D19) |
| `.e2[].precision.bridgedExactRegisteredBand[]` | number | bridged band with exact offset bounds (D19) |
| `.e2[].precision.distanceExactRegisteredBandPp` | number | signed distance on that band, pp |
| `.e2[].precision.cycleBand[]` | number | band rebuilt from exact single-cycle rates |
| `.e2[].precision.bridgedExactCycleBand[]` | number | that band bridged with exact bounds |
| `.e2[].precision.distanceExactCycleBandPp` | number | signed distance on that band, pp |
| `.e2[].precision.movementUnderExactCycleBand` | boolean | verdict under the exact-cycle band (robustness; never replaces the registered verdict) |
| `.e2[].precision.nearerEdge` | number | the band edge nearer the after rate |
| `.e2[].precision.clusterCI95ContainsNearerEdge` | boolean | true marks a borderline verdict (D19) |
| `.e2[].attribution.e1Answers` | number | after-window answers of the engine that qualify under E1 |
| `.e2[].attribution.e1Named` | number | of those, answers naming the medicine |
| `.e2[].attribution.otherAnswers` | number | the engine's other after-window answers |
| `.e2[].attribution.otherNamed` | number | of those, answers naming the medicine |
| `.e2[].attribution.fisherP` | number or null | two-sided Fisher exact p (I3); null when E1 = 0 |
| `.e2[].attribution.verdict` | string | congress effect \| movement without congress sources \| congress sources without movement \| neither (§4) |
| `.e2[].attribution` | null | null where no attribution applies |
| `.bridgeSelfCheck.registered` | string | the §6 sentence registering the check |
| `.bridgeSelfCheck.note` | string | how it is reported (D23) |
| `.bridgeSelfCheck.rows[].series` | string | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `.bridgeSelfCheck.rows[].engine` | string | engine: perplexity (Perplexity, sonar) \| google (Gemini, gemini-2.5-flash) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5) |
| `.bridgeSelfCheck.rows[].v4Runs[]` | string | k/n of each v4-setting bridge run |
| `.bridgeSelfCheck.rows[].v5Runs[]` | string | k/n of each v5-setting run |
| `.bridgeSelfCheck.rows[].v4Spread.min` | number | lowest v4 run rate |
| `.bridgeSelfCheck.rows[].v4Spread.max` | number | highest v4 run rate |
| `.bridgeSelfCheck.rows[].v4Spread.widthPp` | number | max − min, pp |
| `.bridgeSelfCheck.rows[].v5Spread.min` | number | lowest v5 run rate |
| `.bridgeSelfCheck.rows[].v5Spread.max` | number | highest v5 run rate |
| `.bridgeSelfCheck.rows[].v5Spread.widthPp` | number | max − min, pp |
| `.bridgeSelfCheck.rows[].registeredBand[]` | number | registered band, proportion |
| `.bridgeSelfCheck.rows[].bandWidthPp` | number | band width, pp |
| `.bridgeSelfCheck.rows[].v4SpreadOverBandWidth` | number or null | v4 spread ÷ band width; null at ceiling |
| `.bridgeSelfCheck.rows[].reproducesBandWidth` | boolean or null | whether the spread reproduces the band width (shown, not scored); null at ceiling |
| `.singaporePositioning[].slug` | string | cycle code: UAE-Cnn \| SG-Cnn \| SA-Cnn, a neutral identifier of one scan of one series' battery on four engines (the study's internal cycle identifiers are withheld, R5a; fields keep their frozen names, such as slug and cycleSlug) |
| `.singaporePositioning[].date` | string | UTC date |
| `.singaporePositioning[].window` | string | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `.singaporePositioning[].avgRank` | number | governed average rank of Jardiance in answers (1 = named first); descriptive (I4) |
| `.singaporePositioning[].competitors[].name` | string | competitor as configured in the scan |
| `.singaporePositioning[].competitors[].avgRank` | number | its governed average rank |
| `.singaporePositioning[].competitors[].mentionRate` | number | share of governed answers naming it, proportion |
| `.e3.namingsByWindow.{window}.{engine}.namings` | number | E3 (UAE): answers naming an in-scope investigational agent |
| `.e3.namingsByWindow.{window}.{engine}.answers` | number | answers searched |
| `.e3.namingsByWindow.{window}.{engine}.rate` | number | namings/answers |
| `.e3.codesByWindow.{window}.{key}` | number | E3 code counts: status-correct \| status-absent \| status-wrong |
| `.e3.clinicallyMaterialByWindow.{window}` | number | clinically material count: status absent or wrong in a which-medicine answer (§4) |
| `.e3.strictSensitivity.text` | string | definition of the strict scenario (robustness, D16) |
| `.e3.strictSensitivity.materialByWindow.{window}` | number | clinically material count under the strict scenario |
| `.e4.{window}.answers` | number | E4 (quarantined, D17): UAE answers searched |
| `.e4.{window}.withFigure` | number | answers with at least one coded figure |
| `.e4.{window}.byOutcome.{key}` | number | answer-level outcome counts: label \| published \| no-source \| congress (not published as findings; D17) |
| `.e5.{window}.answers` | number | E5: Singapore answers searched |
| `.e5.{window}.candidates` | number | candidate answers found by the term search |
| `.e5.{window}.byCode.{key}` | number | code counts: e5-first-line-general \| e5-first-line-high-risk \| e5-first-line-with-metformin (D2) \| e5-not-first-line |
| `.e5.{window}.protocolFirstLine` | number | registered E5 count: general + high-risk |
| `.e5.{window}.protocolFirstLineStrict` | number | strict scenario (D16) |
| `.saudi.postRegistration` | boolean | true: the Saudi arm was added after registration (D10) |
| `.saudi.e3NamingsByWindow.{window}.{engine}.namings` | number | Saudi E3: answers naming one of the ten agents |
| `.saudi.e3NamingsByWindow.{window}.{engine}.answers` | number | answers searched |
| `.saudi.e3NamingsByWindow.{window}.{engine}.rate` | number or null | namings/answers; null when no answers |
| `.saudi.opportunity.{window}.answers` | number | answers searched (every captured answer of admitted cycles; D21) |
| `.saudi.opportunity.{window}.governedAnswers` | number | governed E2 population, for reference only |
| `.saudi.opportunity.{window}.opportunityAnswers` | number | answers to questions inviting new or emerging treatments (robustness, D16) |
| `.saudi.opportunityRule` | string | the word-list rule applied to question text (regex source) |
| `.saudi.nonProgrammeContext.{key}` | number | descriptive counts of agents outside the programme (never coded; T8) |
| `.saudi.e3iExploratory.{window}.namings` | number | E3-i namings (answer × molecule) |
| `.saudi.e3iExploratory.{window}.sampled` | number | namings coded (cap 60 per molecule per window, seed 20261005) |
| `.saudi.e3iExploratory.{window}.distinctAnswersNaming` | number | distinct answers among the namings |
| `.saudi.e3iExploratory.{window}.t1dUseNamings` | number | coded namings offering the molecule for a type 1 use (denominator, D16) |
| `.saudi.e3iExploratory.{window}.t1dUseDistinctAnswers` | number | distinct answers among them |
| `.saudi.e3iExploratory.{window}.noStatusStatement` | number | of those, indication absent or wrong |
| `.saudi.e3iExploratory.{window}.share` | number or null | noStatusStatement ÷ t1dUseNamings; null when 0 |
| `.saudi.e3iExploratory.{window}.clusterCI95[]` | number | question-clustered 95% interval |
| `.saudi.e3iExploratory.{window}.questions` | number | question clusters |
| `.saudi.e3iExploratory.{window}.inWhichMedicineAnswers` | number | of noStatusStatement, those in which-medicine answers |
| `.saudi.e3iExploratory.{window}.strictShare` | number or null | share under the strict scenario |
| `.saudi.e3iExploratory.{window}.clusterCI95` | null | null when no namings |

## `analysis/easd-2026-analysis.md`

- **Role:** analysis-md. The human-readable tables printed by the same run as the analysis JSON.
- **Endpoint:** As the analysis JSON
- **Source:** `docs/evidence/easd-2026-analysis.md` (frozen); transformed (see REDACTION-LOG.md)
- **Structure:** tables for E2 (UAE, with the full-precision table), Singapore positioning, E3, E4 (quarantined), E5 and the Saudi arm, printed from the analysis JSON by the same run.

## `bridge/easd-2026-bridge.json`

- **Role:** bridge. Per-run naming counts at the v4 and v5 settings, offsets with bootstrap intervals, negative control.
- **Endpoint:** Protocol §6 instrument bridge (v4 to v5)
- **Source:** `docs/evidence/computed/easd-2026-bridge.json` (frozen); transformed (see REDACTION-LOG.md)

| Field | Type | Meaning, units and allowed values |
|---|---|---|
| `.generatedBy` | string | provenance string (script, protocol section) |
| `.protocol` | string | provenance string (script, protocol section) |
| `.design.runsPerArm` | number | R = 3 |
| `.design.order[]` | string | arm order A-B-B-A-A-B (v4 \| v5) |
| `.design.bootstrap` | number | bootstrap resamples (2,000; seed 20261005) |
| `.design.population` | string | governed population (focused questions excluded) |
| `.series.{series}.registered` | boolean | registered arm |
| `.series.{series}.sourceCycle` | string | the admitted v5 cycle whose battery was re-asked |
| `.series.{series}.sourceScannedAt` | string | its completion time, UTC |
| `.series.{series}.engineVersionAtRun` | number | instrument version when the bridge ran |
| `.series.{series}.ranOn` | string | UTC date |
| `.series.{series}.firstOutcomeAt` | string | first outcome, UTC |
| `.series.{series}.lastOutcomeAt` | string | last outcome, UTC |
| `.series.{series}.validation.matchesGovernedCounts` | boolean | the scorer reproduced the source cycle's governed counts before any paid call |
| `.series.{series}.validation.rescored.{engine}.k` | number | re-scored naming count |
| `.series.{series}.validation.rescored.{engine}.n` | number | re-scored answers |
| `.series.{series}.questions` | number | questions in the governed battery |
| `.series.{series}.failedCalls` | number | engine calls that failed (excluded, never read as not named) |
| `.series.{series}.runs[].arm` | string | v4 \| v5 |
| `.series.{series}.runs[].run` | number | position in the order, 1–6 |
| `.series.{series}.runs[].byEngine.{engine}.k` | number | answers naming the medicine |
| `.series.{series}.runs[].byEngine.{engine}.n` | number | answers scored |
| `.series.{series}.offsetsV5minusV4.{engine}.pairedQuestions` | number | question × engine pairs |
| `.series.{series}.offsetsV5minusV4.{engine}.offset` | number | mean v5 − v4 naming difference, proportion |
| `.series.{series}.offsetsV5minusV4.{engine}.ci95[]` | number | bootstrap 95% interval |
| `.series.{series}.negativeControl.engine` | string | openai (ChatGPT, unchanged path) |
| `.series.{series}.negativeControl.holds` | boolean | its interval contains 0 |
| `.series.{series}.verdict` | string | VALID \| FAILED |
| `.series.{series}.meteredCostUsd` | number | metered cost of the completing process, USD (D21 item 5) |

## `code/lib.ts`

- **Role:** script-lib. Study script, sanitised.
- **Endpoint:** Study pipeline
- **Source:** `scripts/easd-2026-study/lib.ts` (frozen); transformed (see REDACTION-LOG.md)
- **Structure:** source code; its header comment states what it reads, writes and never does, and its usage.

## `code/verify.ts`

- **Role:** script-verify. Study script, sanitised.
- **Endpoint:** Study pipeline
- **Source:** `scripts/easd-2026-study/verify.ts` (frozen); transformed (see REDACTION-LOG.md)
- **Structure:** source code; its header comment states what it reads, writes and never does, and its usage.

## `coding/baseline/easd-2026-baseline-coding.json`

- **Role:** baseline-coding. Baseline coding of the registered arms, produced before any after-window cycle existed (D0).
- **Endpoint:** E3, E4, E5 baseline (UAE, Singapore)
- **Source:** `docs/evidence/computed/easd-2026-baseline-coding.json` (frozen); transformed (see REDACTION-LOG.md)

| Field | Type | Meaning, units and allowed values |
|---|---|---|
| `.deterministic` | boolean | true: no wall-clock field |
| `.generatedBy` | string | provenance string (script, protocol section) |
| `.protocol` | string | provenance string (script, protocol section) |
| `.phase` | string | baseline (pre-congress cutoff) \| full (every admitted cycle) |
| `.inputs.pull.pulledAt` | string | pull time, UTC |
| `.inputs.pull.from` | string | first date |
| `.inputs.pull.to` | string | last date |
| `.inputs.statusFetchedAt` | string | time of the regulator snapshot used |
| `.inputs.refs` | string | E4 reference file |
| `.inputs.manual` | string | hand-code file |
| `.inputs.manualCoder` | string | who hand-coded |
| `.inputs.manualCodedAt` | string | date |
| `.schemeDecisions[].id` | string | S1–S11 (registered arms) or T1–T9 (Saudi arm) |
| `.schemeDecisions[].text` | string | the decision, verbatim |
| `.admissions[].series` | string | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `.admissions[].slug` | string | cycle code: UAE-Cnn \| SG-Cnn \| SA-Cnn, a neutral identifier of one scan of one series' battery on four engines (the study's internal cycle identifiers are withheld, R5a; fields keep their frozen names, such as slug and cycleSlug) |
| `.admissions[].date` | string | UTC date |
| `.admissions[].window` | string | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `.admissions[].engineVersion` | number | 4 \| 5 |
| `.admissions[].admitted` | boolean | passes the §2 admission rule |
| `.admissions[].reasons[]` | (empty) or string | reasons for exclusion (empty when admitted) |
| `.admissions[].promptsCaptured` | number | questions with captures |
| `.admissions[].promptCount` | number | questions in the battery |
| `.admissions[].promptsMatchingFirst` | number | questions identical to the series' first cycle |
| `.admissions[].captures` | number | answer captures |
| `.answersPerEngineCycle.{key}` | number | answers captured (key: [series\|]cycle\|engine) |
| `.headline.e3.namingAnswersInScope` | number | E3 naming answers in scope |
| `.headline.e3.byCode.{key}` | number | counts by code (status-correct \| status-absent \| status-wrong \| UNCODED) |
| `.headline.e3.inChoiceQuestions` | number | naming answers to which-medicine questions |
| `.headline.e3.clinicallyMaterial_absentOrWrong_inChoice` | number | clinically material count (§4) |
| `.headline.e3.byAgent.{key}` | number | namings by agent |
| `.headline.e3.outOfScopeContext.{key}` | number | namings outside scope (key: series\|agent; for example an agent approved in the market) |
| `.headline.e4.answersWithCodedFigure` | number | E4 answers with a coded figure |
| `.headline.e4.byOutcome.{key}` | number | label \| published \| no-source \| congress (answer level) |
| `.headline.e4.figures` | number | figures coded |
| `.headline.e4.figuresByCategory.{key}` | number | figures by category |
| `.headline.e4.figuresByPrecision.{key}` | number | decimal \| integer \| converted |
| `.headline.e4.decimalFiguresByCategory.{key}` | number | decimal figures by category |
| `.headline.e4.figuresExcludedByRule` | number | figures excluded by rule (S5) |
| `.headline.e5.candidates` | number | E5 candidates |
| `.headline.e5.byCode.{key}` | number | code counts |
| `.headline.e5.protocolFirstLine_generalPlusHighRisk` | number | registered E5 count |
| `.e3.rows[].series` | string | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `.e3.rows[].slug` | string | cycle code: UAE-Cnn \| SG-Cnn \| SA-Cnn, a neutral identifier of one scan of one series' battery on four engines (the study's internal cycle identifiers are withheld, R5a; fields keep their frozen names, such as slug and cycleSlug) |
| `.e3.rows[].date` | string | UTC date |
| `.e3.rows[].window` | string | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `.e3.rows[].engineVersion` | number | 4 \| 5 |
| `.e3.rows[].engine` | string | engine: perplexity (Perplexity, sonar) \| google (Gemini, gemini-2.5-flash) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5) |
| `.e3.rows[].promptId` | string | opaque question identifier (R2e): <uae\|sg\|sa>-<f\|q><NN>; f = focused question, q = brand-neutral |
| `.e3.rows[].choiceQuestion` | boolean | which-medicine question (S4) |
| `.e3.rows[].agent` | string | retatrutide \| cagrisema \| orforglipron |
| `.e3.rows[].marketStatus` | string | regulator status on the day, as recorded |
| `.e3.rows[].provisional` | boolean | code provisional (status not established) |
| `.e3.rows[].code` | string | status-correct \| status-absent \| status-wrong |
| `.e3.rows[].span` | string | qualifying answer span; on Gemini rows the marker [Gemini answer text withheld] |
| `.e3.rows[].spanSha256` | string | sha256 of the withheld Gemini answer span (present only where the span is withheld, R1m; the Gemini reporting policy) |
| `.e3.rows[].note` | string | coder's note |
| `.e3.rows[].sha256` | string | sha256 of the stored answer capture; links rows across files and to the held pull |
| `.e3.perEngineCycle.{cycle}.{engine}.{key}` | number | code counts per cycle and engine |
| `.e3.outOfScopeContextPerEngineCycle.{key}.{key}` | number | out-of-scope namings (key: series\|agent, then cycle\|engine) |
| `.e4.rows[].slug` | string | cycle code: UAE-Cnn \| SG-Cnn \| SA-Cnn, a neutral identifier of one scan of one series' battery on four engines (the study's internal cycle identifiers are withheld, R5a; fields keep their frozen names, such as slug and cycleSlug) |
| `.e4.rows[].date` | string | UTC date |
| `.e4.rows[].window` | string | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `.e4.rows[].engineVersion` | number | 4 \| 5 |
| `.e4.rows[].engine` | string | engine: perplexity (Perplexity, sonar) \| google (Gemini, gemini-2.5-flash) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5) |
| `.e4.rows[].promptId` | string | opaque question identifier (R2e) |
| `.e4.rows[].choiceQuestion` | boolean | which-medicine question |
| `.e4.rows[].outcome` | string | label \| published \| no-source \| congress (answer level; quarantined, D17) |
| `.e4.rows[].figures[].text` | string | figure as written; a marker on Gemini rows (R1m) and beside a P61 sentence (R1k) |
| `.e4.rows[].figures[].value` | number | numeric value; null beside a P61 sentence (R1k) |
| `.e4.rows[].figures[].decimals` | number | decimal places stated; null beside a P61 sentence |
| `.e4.rows[].figures[].unit` | string | % \| kg \| lb; null beside a P61 sentence |
| `.e4.rows[].figures[].measure` | string | weight-pct \| weight-kg \| hba1c-pp |
| `.e4.rows[].figures[].category` | string | label \| published \| no-source \| congress |
| `.e4.rows[].figures[].matchedRefs[]` | string | reference value ids matched |
| `.e4.rows[].figures[].precision` | string | decimal \| integer \| converted |
| `.e4.rows[].figures[].span` | string | context span; a marker on Gemini rows |
| `.e4.rows[].sha256` | string | sha256 of the stored answer capture; links rows across files and to the held pull |
| `.e4.rows[].figures[].textSha256` | string | sha256 of the withheld Gemini figure string (present only where it is withheld, R1m) |
| `.e4.rows[].figures[].spanSha256` | string | sha256 of the withheld Gemini answer span (present only where the span is withheld, R1m; the Gemini reporting policy) |
| `.e4.perEngineCycle.{cycle}.{engine}.{key}` | number | outcome counts |
| `.e5.rows[].slug` | string | cycle code: UAE-Cnn \| SG-Cnn \| SA-Cnn, a neutral identifier of one scan of one series' battery on four engines (the study's internal cycle identifiers are withheld, R5a; fields keep their frozen names, such as slug and cycleSlug) |
| `.e5.rows[].date` | string | UTC date |
| `.e5.rows[].window` | string | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `.e5.rows[].engineVersion` | number | 4 \| 5 |
| `.e5.rows[].engine` | string | engine: perplexity (Perplexity, sonar) \| google (Gemini, gemini-2.5-flash) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5) |
| `.e5.rows[].promptId` | string | opaque question identifier (R2e) |
| `.e5.rows[].choiceQuestion` | boolean | which-medicine question |
| `.e5.rows[].terms[]` | string | candidate terms found (first-line \| initial \| initially \| monotherapy …) |
| `.e5.rows[].code` | string | E5 code |
| `.e5.rows[].span` | string | qualifying span; a marker on Gemini rows |
| `.e5.rows[].note` | string | note |
| `.e5.rows[].sha256` | string | sha256 of the stored answer capture; links rows across files and to the held pull |
| `.e5.rows[].spanSha256` | string | sha256 of the withheld Gemini answer span (present only where the span is withheld, R1m; the Gemini reporting policy) |
| `.e5.perEngineCycle.{cycle}.{engine}.{key}` | number | code counts |

## `coding/baseline/easd-2026-coding-ledger.md`

- **Role:** baseline-coding-ledger. The baseline coding ledger: scheme decisions S1–S11 and every coded row.
- **Endpoint:** E3, E4, E5 baseline
- **Source:** `docs/evidence/easd-2026-coding-ledger.md` (committed, not frozen); transformed (see REDACTION-LOG.md)
- **Structure:** scheme decisions S1–S11, admission table, then one table per endpoint (E3, E4, E5) with one row per coded answer or figure: cycle, date, instrument version, engine, opaque question id, which-medicine flag, code or outcome, span and note.

## `coding/baseline/easd-2026-saudi-baseline-coding.json`

- **Role:** saudi-baseline-coding. Baseline coding of the Saudi arm with a 26 Sep cutoff.
- **Endpoint:** Saudi E3 and E3-i baseline (post-registration, D10)
- **Source:** `docs/evidence/computed/easd-2026-saudi-baseline-coding.json` (frozen); transformed (see REDACTION-LOG.md)

| Field | Type | Meaning, units and allowed values |
|---|---|---|
| `.deterministic` | boolean | true: no wall-clock field |
| `.generatedBy` | string | provenance string (script, protocol section) |
| `.protocol` | string | provenance string (script, protocol section) |
| `.series.key` | string | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `.series.brand` | string | measured medicine |
| `.series.market` | string | regulator market |
| `.series.registered` | boolean | false for the Saudi arm |
| `.phase` | string | baseline (pre-congress cutoff) \| full (every admitted cycle) |
| `.inputs.pull.pulledAt` | string | pull time, UTC |
| `.inputs.pull.from` | string | first date |
| `.inputs.pull.to` | string | last date |
| `.inputs.statusFetchedAt` | string | time of the regulator snapshot used |
| `.inputs.manual` | string | hand-code file |
| `.inputs.manualCoder` | string | who hand-coded |
| `.inputs.manualCodedAt` | string | date |
| `.schemeDecisions[].id` | string | S1–S11 (registered arms) or T1–T9 (Saudi arm) |
| `.schemeDecisions[].text` | string | the decision, verbatim |
| `.admissions[].slug` | string | cycle code: UAE-Cnn \| SG-Cnn \| SA-Cnn, a neutral identifier of one scan of one series' battery on four engines (the study's internal cycle identifiers are withheld, R5a; fields keep their frozen names, such as slug and cycleSlug) |
| `.admissions[].date` | string | UTC date |
| `.admissions[].window` | string | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `.admissions[].engineVersion` | number | 4 \| 5 |
| `.admissions[].admitted` | boolean | passes the §2 admission rule |
| `.admissions[].reasons[]` | (empty) | reasons for exclusion (empty when admitted) |
| `.admissions[].promptsCaptured` | number | questions with captures |
| `.admissions[].promptCount` | number | questions in the battery |
| `.admissions[].promptsMatchingFirst` | number | questions identical to the series' first cycle |
| `.admissions[].captures` | number | answer captures |
| `.answersPerEngineCycle.{key}` | number | answers captured (key: [series\|]cycle\|engine) |
| `.headline.label` | string | post-registration label |
| `.headline.e3.namingAnswersInScope` | number | E3 naming answers in scope |
| `.headline.e3.removedNotTheAgent` | number | Saudi: matches removed as not the agent |
| `.headline.e3.byCode.{key}` | number | counts by code (status-correct \| status-absent \| status-wrong \| UNCODED) |
| `.headline.e3.inChoiceQuestions` | number | naming answers to which-medicine questions |
| `.headline.e3.clinicallyMaterial_absentOrWrong_inChoice` | number | clinically material count (§4) |
| `.headline.e3.allCodesProvisional` | boolean | Saudi: every code provisional (status not established, D12) |
| `.headline.e3.byAgent.{key}` | number | namings by agent |
| `.headline.e3i.exploratory` | boolean | true (D12) |
| `.headline.e3i.namings` | number | E3-i namings (answer × molecule) |
| `.headline.e3i.distinctAnswers` | number | distinct answers |
| `.headline.e3i.byAgent.{key}` | number | namings by molecule |
| `.headline.e3i.byWindowAgent.{window}.{key}.named` | number | namings |
| `.headline.e3i.byWindowAgent.{window}.{key}.sampled` | number | namings coded (cap 60) |
| `.headline.e3i.byCode.{key}` | number | indication-stated \| indication-absent \| indication-wrong \| not-a-t1d-use \| not-the-agent \| UNCODED \| NOT-SAMPLED |
| `.headline.e3i.scoredT1dNamings` | number | namings offered for a type 1 use |
| `.headline.e3i.scoredT1dDistinctAnswers` | number | distinct answers among them |
| `.headline.e3i.indicationNotStated_absentOrWrong.k` | number | absent or wrong |
| `.headline.e3i.indicationNotStated_absentOrWrong.n` | number | type 1 use namings |
| `.headline.e3i.indicationNotStated_absentOrWrong.share` | number | k/n |
| `.headline.e3i.indicationNotStated_absentOrWrong.wilson95[]` | number | Wilson 95% interval |
| `.headline.e3i.exploratoryMaterial_absentOrWrong_inChoice` | number | absent or wrong in which-medicine answers |
| `.headline.nonProgrammeContext.{key}` | number | descriptive counts (T8) |
| `.e3.rows[]` | (empty) | one row per E3 naming (empty for the Saudi arm) |
| `.e3i.rows[].slug` | string | cycle code: UAE-Cnn \| SG-Cnn \| SA-Cnn, a neutral identifier of one scan of one series' battery on four engines (the study's internal cycle identifiers are withheld, R5a; fields keep their frozen names, such as slug and cycleSlug) |
| `.e3i.rows[].date` | string | UTC date |
| `.e3i.rows[].window` | string | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `.e3i.rows[].engineVersion` | number | 4 \| 5 |
| `.e3i.rows[].engine` | string | engine: perplexity (Perplexity, sonar) \| google (Gemini, gemini-2.5-flash) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5) |
| `.e3i.rows[].promptId` | string | opaque question identifier (R2e) |
| `.e3i.rows[].choiceQuestion` | boolean | which-medicine question |
| `.e3i.rows[].agent` | string | sotagliflozin \| tirzepatide \| semaglutide \| bcg \| fenofibrate |
| `.e3i.rows[].sampleRank` | number | rank in the seeded sample (−1 if not drawn) |
| `.e3i.rows[].sampled` | boolean | coded under the cap |
| `.e3i.rows[].code` | string | E3-i code |
| `.e3i.rows[].span` | string | qualifying span; on Gemini rows a marker |
| `.e3i.rows[].note` | string | note |
| `.e3i.rows[].sha256` | string | sha256 of the stored answer capture; links rows across files and to the held pull |
| `.e3i.rows[].spanSha256` | string | sha256 of the withheld Gemini answer span (present only where the span is withheld, R1m; the Gemini reporting policy) |
| `.e3i.perEngineCycle.{cycle}.{engine}.{key}` | number | code counts |
| `.nonProgrammeContextPerEngineCycle.{key}.{cycle}.{engine}` | number | descriptive counts per cycle and engine |

## `coding/baseline/easd-2026-saudi-coding-ledger.md`

- **Role:** saudi-baseline-coding-ledger. The Saudi baseline ledger: scheme decisions T1–T9 and every coded row.
- **Endpoint:** Saudi E3 and E3-i baseline
- **Source:** `docs/evidence/easd-2026-saudi-coding-ledger.md` (committed, not frozen); transformed (see REDACTION-LOG.md)
- **Structure:** scheme decisions T1–T9, admission table, E3 and E3-i rows (E3-i sampled rows only) for the Saudi arm, labelled post-registration.

## `coding/final/easd-2026-final-coding-ledger.md`

- **Role:** final-coding-ledger. The final coding ledger, one row per coded answer or figure.
- **Endpoint:** E3, E4, E5
- **Source:** `docs/evidence/easd-2026-final-coding-ledger.md` (frozen); transformed (see REDACTION-LOG.md)
- **Structure:** as the baseline ledger, over every admitted cycle, with the D4 hand-check results for E4.

## `coding/final/easd-2026-final-coding.json`

- **Role:** final-coding. Final coding of the registered arms over every admitted cycle.
- **Endpoint:** E3, E4, E5 (registered arms), full study period
- **Source:** `docs/evidence/computed/easd-2026-final-coding.json` (frozen); transformed (see REDACTION-LOG.md)

| Field | Type | Meaning, units and allowed values |
|---|---|---|
| `.deterministic` | boolean | true: no wall-clock field |
| `.generatedBy` | string | provenance string (script, protocol section) |
| `.protocol` | string | provenance string (script, protocol section) |
| `.phase` | string | baseline (pre-congress cutoff) \| full (every admitted cycle) |
| `.inputs.pull.pulledAt` | string | pull time, UTC |
| `.inputs.pull.from` | string | first date |
| `.inputs.pull.to` | string | last date |
| `.inputs.statusFetchedAt` | string | time of the regulator snapshot used |
| `.inputs.refs` | string | E4 reference file |
| `.inputs.manual` | string | hand-code file |
| `.inputs.manualCoder` | string | who hand-coded |
| `.inputs.manualCodedAt` | string | date |
| `.schemeDecisions[].id` | string | S1–S11 (registered arms) or T1–T9 (Saudi arm) |
| `.schemeDecisions[].text` | string | the decision, verbatim |
| `.admissions[].series` | string | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `.admissions[].slug` | string | cycle code: UAE-Cnn \| SG-Cnn \| SA-Cnn, a neutral identifier of one scan of one series' battery on four engines (the study's internal cycle identifiers are withheld, R5a; fields keep their frozen names, such as slug and cycleSlug) |
| `.admissions[].date` | string | UTC date |
| `.admissions[].window` | string | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `.admissions[].engineVersion` | number | 4 \| 5 |
| `.admissions[].admitted` | boolean | passes the §2 admission rule |
| `.admissions[].reasons[]` | (empty) or string | reasons for exclusion (empty when admitted) |
| `.admissions[].promptsCaptured` | number | questions with captures |
| `.admissions[].promptCount` | number | questions in the battery |
| `.admissions[].promptsMatchingFirst` | number | questions identical to the series' first cycle |
| `.admissions[].captures` | number | answer captures |
| `.answersPerEngineCycle.{key}` | number | answers captured (key: [series\|]cycle\|engine) |
| `.headline.e3.namingAnswersInScope` | number | E3 naming answers in scope |
| `.headline.e3.byCode.{key}` | number | counts by code (status-correct \| status-absent \| status-wrong \| UNCODED) |
| `.headline.e3.inChoiceQuestions` | number | naming answers to which-medicine questions |
| `.headline.e3.clinicallyMaterial_absentOrWrong_inChoice` | number | clinically material count (§4) |
| `.headline.e3.byAgent.{key}` | number | namings by agent |
| `.headline.e3.outOfScopeContext.{key}` | number | namings outside scope (key: series\|agent; for example an agent approved in the market) |
| `.headline.e4.answersWithCodedFigure` | number | E4 answers with a coded figure |
| `.headline.e4.byOutcome.{key}` | number | label \| published \| no-source \| congress (answer level) |
| `.headline.e4.figures` | number | figures coded |
| `.headline.e4.figuresByCategory.{key}` | number | figures by category |
| `.headline.e4.figuresByPrecision.{key}` | number | decimal \| integer \| converted |
| `.headline.e4.decimalFiguresByCategory.{key}` | number | decimal figures by category |
| `.headline.e4.figuresExcludedByRule` | number | figures excluded by rule (S5) |
| `.headline.e4.d4.unchecked` | number | D4 figures not yet hand-checked (0 in the final file) |
| `.headline.e4.d4.excludedOnReading` | number | excluded on the D4 read |
| `.headline.e4.d4.recategorisedOnReading` | number | recategorised on the D4 read |
| `.headline.e5.candidates` | number | E5 candidates |
| `.headline.e5.byCode.{key}` | number | code counts |
| `.headline.e5.protocolFirstLine_generalPlusHighRisk` | number | registered E5 count |
| `.e3.rows[].series` | string | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `.e3.rows[].slug` | string | cycle code: UAE-Cnn \| SG-Cnn \| SA-Cnn, a neutral identifier of one scan of one series' battery on four engines (the study's internal cycle identifiers are withheld, R5a; fields keep their frozen names, such as slug and cycleSlug) |
| `.e3.rows[].date` | string | UTC date |
| `.e3.rows[].window` | string | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `.e3.rows[].engineVersion` | number | 4 \| 5 |
| `.e3.rows[].engine` | string | engine: perplexity (Perplexity, sonar) \| google (Gemini, gemini-2.5-flash) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5) |
| `.e3.rows[].promptId` | string | opaque question identifier (R2e): <uae\|sg\|sa>-<f\|q><NN>; f = focused question, q = brand-neutral |
| `.e3.rows[].choiceQuestion` | boolean | which-medicine question (S4) |
| `.e3.rows[].agent` | string | retatrutide \| cagrisema \| orforglipron |
| `.e3.rows[].marketStatus` | string | regulator status on the day, as recorded |
| `.e3.rows[].provisional` | boolean | code provisional (status not established) |
| `.e3.rows[].code` | string | status-correct \| status-absent \| status-wrong |
| `.e3.rows[].span` | string | qualifying answer span; on Gemini rows the marker [Gemini answer text withheld] |
| `.e3.rows[].spanSha256` | string | sha256 of the withheld Gemini answer span (present only where the span is withheld, R1m; the Gemini reporting policy) |
| `.e3.rows[].note` | string | coder's note |
| `.e3.rows[].sha256` | string | sha256 of the stored answer capture; links rows across files and to the held pull |
| `.e3.perEngineCycle.{cycle}.{engine}.{key}` | number | code counts per cycle and engine |
| `.e3.outOfScopeContextPerEngineCycle.{key}.{key}` | number | out-of-scope namings (key: series\|agent, then cycle\|engine) |
| `.e4.rows[].slug` | string | cycle code: UAE-Cnn \| SG-Cnn \| SA-Cnn, a neutral identifier of one scan of one series' battery on four engines (the study's internal cycle identifiers are withheld, R5a; fields keep their frozen names, such as slug and cycleSlug) |
| `.e4.rows[].date` | string | UTC date |
| `.e4.rows[].window` | string | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `.e4.rows[].engineVersion` | number | 4 \| 5 |
| `.e4.rows[].engine` | string | engine: perplexity (Perplexity, sonar) \| google (Gemini, gemini-2.5-flash) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5) |
| `.e4.rows[].promptId` | string | opaque question identifier (R2e) |
| `.e4.rows[].choiceQuestion` | boolean | which-medicine question |
| `.e4.rows[].outcome` | string | label \| published \| no-source \| congress (answer level; quarantined, D17) |
| `.e4.rows[].figures[].text` | string | figure as written; a marker on Gemini rows (R1m) and beside a P61 sentence (R1k) |
| `.e4.rows[].figures[].value` | number or null | numeric value; null beside a P61 sentence (R1k) |
| `.e4.rows[].figures[].decimals` | number or null | decimal places stated; null beside a P61 sentence |
| `.e4.rows[].figures[].unit` | string or null | % \| kg \| lb; null beside a P61 sentence |
| `.e4.rows[].figures[].measure` | string | weight-pct \| weight-kg \| hba1c-pp |
| `.e4.rows[].figures[].category` | string | label \| published \| no-source \| congress |
| `.e4.rows[].figures[].matchedRefs[]` | string | reference value ids matched |
| `.e4.rows[].figures[].precision` | string | decimal \| integer \| converted |
| `.e4.rows[].figures[].span` | string | context span; a marker on Gemini rows |
| `.e4.rows[].sha256` | string | sha256 of the stored answer capture; links rows across files and to the held pull |
| `.e4.rows[].figures[].textSha256` | string | sha256 of the withheld Gemini figure string (present only where it is withheld, R1m) |
| `.e4.rows[].figures[].spanSha256` | string | sha256 of the withheld Gemini answer span (present only where the span is withheld, R1m; the Gemini reporting policy) |
| `.e4.rows[].figures[].d4.verdict` | string | confirmed (D4 hand read) |
| `.e4.rows[].figures[].d4.scriptedCategory` | string | category the script gave |
| `.e4.rows[].figures[].d4.note` | string | reader's note |
| `.e4.rows[].figures[].matchedRefs` | null | null beside a P61 sentence (R1k: a reference value would give the figure back) |
| `.e4.perEngineCycle.{cycle}.{engine}.{key}` | number | outcome counts |
| `.e5.rows[].slug` | string | cycle code: UAE-Cnn \| SG-Cnn \| SA-Cnn, a neutral identifier of one scan of one series' battery on four engines (the study's internal cycle identifiers are withheld, R5a; fields keep their frozen names, such as slug and cycleSlug) |
| `.e5.rows[].date` | string | UTC date |
| `.e5.rows[].window` | string | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `.e5.rows[].engineVersion` | number | 4 \| 5 |
| `.e5.rows[].engine` | string | engine: perplexity (Perplexity, sonar) \| google (Gemini, gemini-2.5-flash) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5) |
| `.e5.rows[].promptId` | string | opaque question identifier (R2e) |
| `.e5.rows[].choiceQuestion` | boolean | which-medicine question |
| `.e5.rows[].terms[]` | string | candidate terms found (first-line \| initial \| initially \| monotherapy …) |
| `.e5.rows[].code` | string | E5 code |
| `.e5.rows[].span` | string | qualifying span; a marker on Gemini rows |
| `.e5.rows[].note` | string | note |
| `.e5.rows[].sha256` | string | sha256 of the stored answer capture; links rows across files and to the held pull |
| `.e5.rows[].spanSha256` | string | sha256 of the withheld Gemini answer span (present only where the span is withheld, R1m; the Gemini reporting policy) |
| `.e5.perEngineCycle.{cycle}.{engine}.{key}` | number | code counts |

## `coding/final/easd-2026-saudi-final-coding-ledger.md`

- **Role:** saudi-final-coding-ledger. The Saudi final coding ledger.
- **Endpoint:** Saudi E3 and E3-i
- **Source:** `docs/evidence/easd-2026-saudi-final-coding-ledger.md` (frozen); transformed (see REDACTION-LOG.md)
- **Structure:** as the Saudi baseline ledger, over every admitted cycle.

## `coding/final/easd-2026-saudi-final-coding.json`

- **Role:** saudi-final-coding. Final coding of the Saudi arm.
- **Endpoint:** Saudi E3 and E3-i (post-registration)
- **Source:** `docs/evidence/computed/easd-2026-saudi-final-coding.json` (frozen); transformed (see REDACTION-LOG.md)

| Field | Type | Meaning, units and allowed values |
|---|---|---|
| `.deterministic` | boolean | true: no wall-clock field |
| `.generatedBy` | string | provenance string (script, protocol section) |
| `.protocol` | string | provenance string (script, protocol section) |
| `.series.key` | string | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `.series.brand` | string | measured medicine |
| `.series.market` | string | regulator market |
| `.series.registered` | boolean | false for the Saudi arm |
| `.phase` | string | baseline (pre-congress cutoff) \| full (every admitted cycle) |
| `.inputs.pull.pulledAt` | string | pull time, UTC |
| `.inputs.pull.from` | string | first date |
| `.inputs.pull.to` | string | last date |
| `.inputs.statusFetchedAt` | string | time of the regulator snapshot used |
| `.inputs.manual` | string | hand-code file |
| `.inputs.manualCoder` | string | who hand-coded |
| `.inputs.manualCodedAt` | string | date |
| `.schemeDecisions[].id` | string | S1–S11 (registered arms) or T1–T9 (Saudi arm) |
| `.schemeDecisions[].text` | string | the decision, verbatim |
| `.admissions[].slug` | string | cycle code: UAE-Cnn \| SG-Cnn \| SA-Cnn, a neutral identifier of one scan of one series' battery on four engines (the study's internal cycle identifiers are withheld, R5a; fields keep their frozen names, such as slug and cycleSlug) |
| `.admissions[].date` | string | UTC date |
| `.admissions[].window` | string | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `.admissions[].engineVersion` | number | 4 \| 5 |
| `.admissions[].admitted` | boolean | passes the §2 admission rule |
| `.admissions[].reasons[]` | (empty) or string | reasons for exclusion (empty when admitted) |
| `.admissions[].promptsCaptured` | number | questions with captures |
| `.admissions[].promptCount` | number | questions in the battery |
| `.admissions[].promptsMatchingFirst` | number | questions identical to the series' first cycle |
| `.admissions[].captures` | number | answer captures |
| `.answersPerEngineCycle.{key}` | number | answers captured (key: [series\|]cycle\|engine) |
| `.headline.label` | string | post-registration label |
| `.headline.e3.namingAnswersInScope` | number | E3 naming answers in scope |
| `.headline.e3.removedNotTheAgent` | number | Saudi: matches removed as not the agent |
| `.headline.e3.byCode.{key}` | number | counts by code (status-correct \| status-absent \| status-wrong \| UNCODED) |
| `.headline.e3.inChoiceQuestions` | number | naming answers to which-medicine questions |
| `.headline.e3.clinicallyMaterial_absentOrWrong_inChoice` | number | clinically material count (§4) |
| `.headline.e3.allCodesProvisional` | boolean | Saudi: every code provisional (status not established, D12) |
| `.headline.e3.byAgent.{key}` | number | namings by agent |
| `.headline.e3i.exploratory` | boolean | true (D12) |
| `.headline.e3i.namings` | number | E3-i namings (answer × molecule) |
| `.headline.e3i.distinctAnswers` | number | distinct answers |
| `.headline.e3i.byAgent.{key}` | number | namings by molecule |
| `.headline.e3i.byWindowAgent.{window}.{key}.named` | number | namings |
| `.headline.e3i.byWindowAgent.{window}.{key}.sampled` | number | namings coded (cap 60) |
| `.headline.e3i.byCode.{key}` | number | indication-stated \| indication-absent \| indication-wrong \| not-a-t1d-use \| not-the-agent \| UNCODED \| NOT-SAMPLED |
| `.headline.e3i.scoredT1dNamings` | number | namings offered for a type 1 use |
| `.headline.e3i.scoredT1dDistinctAnswers` | number | distinct answers among them |
| `.headline.e3i.indicationNotStated_absentOrWrong.k` | number | absent or wrong |
| `.headline.e3i.indicationNotStated_absentOrWrong.n` | number | type 1 use namings |
| `.headline.e3i.indicationNotStated_absentOrWrong.share` | number | k/n |
| `.headline.e3i.indicationNotStated_absentOrWrong.wilson95[]` | number | Wilson 95% interval |
| `.headline.e3i.exploratoryMaterial_absentOrWrong_inChoice` | number | absent or wrong in which-medicine answers |
| `.headline.nonProgrammeContext.{key}` | number | descriptive counts (T8) |
| `.e3.rows[]` | (empty) | one row per E3 naming (empty for the Saudi arm) |
| `.e3i.rows[].slug` | string | cycle code: UAE-Cnn \| SG-Cnn \| SA-Cnn, a neutral identifier of one scan of one series' battery on four engines (the study's internal cycle identifiers are withheld, R5a; fields keep their frozen names, such as slug and cycleSlug) |
| `.e3i.rows[].date` | string | UTC date |
| `.e3i.rows[].window` | string | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `.e3i.rows[].engineVersion` | number | 4 \| 5 |
| `.e3i.rows[].engine` | string | engine: perplexity (Perplexity, sonar) \| google (Gemini, gemini-2.5-flash) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5) |
| `.e3i.rows[].promptId` | string | opaque question identifier (R2e) |
| `.e3i.rows[].choiceQuestion` | boolean | which-medicine question |
| `.e3i.rows[].agent` | string | sotagliflozin \| tirzepatide \| semaglutide \| bcg \| fenofibrate |
| `.e3i.rows[].sampleRank` | number | rank in the seeded sample (−1 if not drawn) |
| `.e3i.rows[].sampled` | boolean | coded under the cap |
| `.e3i.rows[].code` | string | E3-i code |
| `.e3i.rows[].span` | string | qualifying span; on Gemini rows a marker |
| `.e3i.rows[].note` | string | note |
| `.e3i.rows[].sha256` | string | sha256 of the stored answer capture; links rows across files and to the held pull |
| `.e3i.rows[].spanSha256` | string | sha256 of the withheld Gemini answer span (present only where the span is withheld, R1m; the Gemini reporting policy) |
| `.e3i.perEngineCycle.{cycle}.{engine}.{key}` | number | code counts |
| `.nonProgrammeContextPerEngineCycle.{key}.{cycle}.{engine}` | number | descriptive counts per cycle and engine |

## `coding/manual/easd-2026-manual-codes.json`

- **Role:** manual-codes. Hand codes and adjudications for the registered arms, keyed by answer hash.
- **Endpoint:** E3, E4 (D4 hand check), E5
- **Source:** `docs/evidence/easd-2026-manual-codes.json` (frozen); transformed (see REDACTION-LOG.md)

| Field | Type | Meaning, units and allowed values |
|---|---|---|
| `.coder` | string | who coded |
| `.codedAt` | string | date |
| `.rules` | string | rules applied |
| `.e3.{sha256}[].agent` | string | agent |
| `.e3.{sha256}[].code` | string | E3 code |
| `.e3.{sha256}[].span` | string | span (a marker for Gemini answers) |
| `.e3.{sha256}[].spanSha256` | string | sha256 of the withheld Gemini answer span (present only where the span is withheld, R1m; the Gemini reporting policy) |
| `.e3.{sha256}[].note` | string | note |
| `.e5.{sha256}.code` | string | E5 code |
| `.e5.{sha256}.span` | string | span (a marker for Gemini answers) |
| `.e5.{sha256}.spanSha256` | string | sha256 of the withheld Gemini answer span (present only where the span is withheld, R1m; the Gemini reporting policy) |
| `.e5.{sha256}.note` | string | note |
| `.e5.{sha256}.firstCoderCode` | string | first coder's code when adjudicated |
| `.e5.{sha256}.adjudication` | string | adjudication |
| `.afterWindowCoder` | string | who coded the after window |
| `.afterWindowCodedThrough.codedAt` | string | date |
| `.afterWindowCodedThrough.cyclesThrough` | string | last cycle date coded |
| `.e4.{sha256}[].text` | string | figure as written (a marker for Gemini answers and beside a P61 sentence) |
| `.e4.{sha256}[].value` | number or null | value; null beside a P61 sentence |
| `.e4.{sha256}[].verdict` | string | confirmed \| excluded (D4) |
| `.e4.{sha256}[].note` | string | note |
| `.e4.{sha256}[].textSha256` | string | sha256 of the withheld Gemini figure string (present only where it is withheld, R1m) |
| `.e4.{sha256}[].reason` | string | reason code (P1–U1) or empty |
| `.e4Checker` | string | who ran the D4 read |
| `.e4ReasonCodes.{key}` | string | reason code definition |

## `coding/manual/easd-2026-saudi-manual-codes.json`

- **Role:** saudi-manual-codes. Hand codes and adjudications for the Saudi arm, keyed by answer hash.
- **Endpoint:** Saudi E3-i
- **Source:** `docs/evidence/easd-2026-saudi-manual-codes.json` (frozen); transformed (see REDACTION-LOG.md)

| Field | Type | Meaning, units and allowed values |
|---|---|---|
| `.coder` | string | who coded |
| `.codedAt` | string | date |
| `.e3i.{sha256}[].agent` | string | molecule |
| `.e3i.{sha256}[].code` | string | E3-i code |
| `.e3i.{sha256}[].span` | string | span (a marker for Gemini answers) |
| `.e3i.{sha256}[].note` | string | note |
| `.e3i.{sha256}[].spanSha256` | string | sha256 of the withheld Gemini answer span (present only where the span is withheld, R1m; the Gemini reporting policy) |
| `.e3i.{sha256}[].firstCoderCode` | string | first code when adjudicated |
| `.e3i.{sha256}[].adjudication` | string | adjudication |
| `.afterWindowCoder` | string | who coded the after window |
| `.afterWindowCodedThrough.codedAt` | string | date |
| `.afterWindowCodedThrough.cyclesThrough` | string | last cycle coded |

## `DATA-DICTIONARY.md`

- **Role:** data-dictionary. Every public file and field.
- **Endpoint:** All
- **Source:** generated by the build; —
- **Structure:** this file.

## `e1/easd-2026-e1-ledger.json`

- **Role:** e1-ledger. The hand ledger: every screened candidate page and text candidate, classified, with the diaTribe adjudication.
- **Endpoint:** E1 source and text routes
- **Source:** `docs/evidence/easd-2026-e1-ledger.json` (frozen); transformed (see REDACTION-LOG.md)

| Field | Type | Meaning, units and allowed values |
|---|---|---|
| `.coder` | string | who classified |
| `.codedAt` | string | date |
| `.scope` | string | cycles covered |
| `.sources[].series` | string | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `.sources[].url` | string | page URL (withheld pages as typed tokens; a page cited only by Gemini as a neutral token, R1o) |
| `.sources[].class` | string | a-congress \| b-journal \| c-sponsor \| d-media \| not-congress-derived (§4 E1) |
| `.sources[].readouts[]` | string | readouts the page carries |
| `.sources[].pageDate` | string or null | article date as printed (YYYY-MM-DD; the month only, YYYY-MM, where the title or path is withheld, R1q); null if none |
| `.sources[].fetchedAt` | string | UTC time the page was read |
| `.sources[].qualifying` | string | the sentence that qualifies the page (empty when excluded) |
| `.sources[].note` | string | the reader's note |
| `.sources[].readVia` | string | browser when read in an ordinary browser; empty when the scripted fetch sufficed |
| `.sources[].revisions[].revisedAt` | string | UTC time of a revision |
| `.sources[].revisions[].previousClass` | string | earlier class |
| `.sources[].revisions[].previousReadouts[]` | string | earlier readouts |
| `.sources[].revisions[].reason` | string | reason |
| `.sources[].readoutsUnverifiedAtCitation[]` | string | readouts not shown to be on the page when cited (D19) |
| `.sources[].adjudication` | string | adjudication, verbatim |
| `.sources[].contentAtCitation` | string | what the page can be shown to have said at citation |
| `.text[].sha256` | string | sha256 of the stored answer capture; links rows across files and to the held pull |
| `.text[].qualifies` | boolean | whether the answer's text reports a readout result |
| `.text[].readouts[]` | (empty) | readouts reported |
| `.text[].span` | string | answer span read (≤ about 240 characters); a marker for Gemini answers and for answers whose engine no public file records |
| `.text[].spanSha256` | string | sha256 of the withheld Gemini answer span (present only where the span is withheld, R1m; the Gemini reporting policy) |
| `.text[].note` | string | reader's note |

## `e1/easd-2026-e1-manifest.json`

- **Role:** e1-manifest. Every new (series, URL) pair cited in the congress and after windows, as cited and resolved, with readability and screen result.
- **Endpoint:** E1 source route (screen)
- **Source:** `docs/evidence/computed/easd-2026-e1-manifest.json` (frozen); transformed (see REDACTION-LOG.md)

| Field | Type | Meaning, units and allowed values |
|---|---|---|
| `.generatedBy` | string | provenance string (script, protocol section) |
| `.pull.pulledAt` | string | UTC time of the pull used (16:36, identical cycles; D21) |
| `.pull.from` | string | first date |
| `.pull.to` | string | last date |
| `.note` | string | what the manifest covers |
| `.entries[].series` | string | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `.entries[].url` | string | resolved URL; withheld pages as [type-NN]/pNNN tokens; a page cited only by Gemini as [gemini-only]/pNNNN, with its publisher and title withheld (R1o, the release policy) |
| `.entries[].citedAs[]` | string | URL as cited; a Gemini grounding redirect as the marker [Gemini grounding redirect] (R1l); withheld forms tokenised |
| `.entries[].engines[]` | string | engine: perplexity (Perplexity, sonar) \| google (Gemini, gemini-2.5-flash) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5) |
| `.entries[].firstObservedInAnswer` | string | UTC capture time of the first answer citing it |
| `.entries[].answers` | number | answers citing it |
| `.entries[].fetchedAt` | string | UTC time of the fetch |
| `.entries[].httpStatus` | string or number | HTTP status or fetch error of the last attempt |
| `.entries[].firstHttpStatus` | string or null or number | status of the first attempt (null when only one) |
| `.entries[].readable` | boolean | true if the page could be read |
| `.entries[].pageDate` | null or string | date found on the page, if any (the month only for page P61 and for pages whose title or path is withheld, R1k and R1q) |
| `.entries[].title` | string | page title (withheld for withheld pages) |
| `.entries[].screened` | string | candidate \| no congress or readout term |

## `e1/easd-2026-e1.json`

- **Role:** e1. E1 tallies per engine and cycle, latency, cross-arm readouts, inspection coverage.
- **Endpoint:** E1 (registered; Saudi arm post-registration)
- **Source:** `docs/evidence/computed/easd-2026-e1.json` (frozen); transformed (see REDACTION-LOG.md)

| Field | Type | Meaning, units and allowed values |
|---|---|---|
| `.deterministic` | boolean | true: no wall-clock field |
| `.generatedBy` | string | provenance string (script, protocol section) |
| `.protocol` | string | provenance string (script, protocol section) |
| `.inputs.pull.pulledAt` | string | UTC time of the pull used |
| `.inputs.pull.from` | string | first date |
| `.inputs.pull.to` | string | last date |
| `.inputs.ledger` | string | ledger path |
| `.inputs.ledgerCoder` | string | who read the candidates |
| `.inputs.ledgerCodedAt` | string | date of the ledger |
| `.inputs.manifest` | string | manifest path |
| `.evidenceNote` | string | how a source's class is dated |
| `.detectionLimit` | string | what the new-URL rule cannot detect |
| `.screen.newPairs` | number | new (series, URL) pairs in the congress and after windows |
| `.screen.uniqueUrls` | number | distinct URLs among them |
| `.screen.candidates` | number | pages carrying a congress or readout term, read by hand |
| `.screen.screenedOut` | number | pages carrying no such term |
| `.screen.unreadable` | number | pages that could not be read (unverified, never negative) |
| `.screen.candidateKeys[]` | string | series\|URL of every hand-read candidate (withheld URLs as tokens) |
| `.readouts[].id` | string | readout label |
| `.readouts[].series[]` | string | series the readout is registered for |
| `.readouts[].presentedUtc` | string | programme presentation time, UTC |
| `.readouts[].postRegistration` | null or string | null if registered; otherwise the deviation or label that added it |
| `.ledgerCounts.sourcesQualifying` | number | ledger sources qualifying under §4 classes (a)–(d) |
| `.ledgerCounts.sourcesExcludedOnReading` | number | candidates excluded on reading |
| `.ledgerCounts.textQualifying` | number | answers whose text reports a readout result |
| `.ledgerCounts.textExcludedOnReading` | number | text candidates excluded |
| `.uncodedTextCandidates` | number | text candidates left unread (0) |
| `.windows.{key}.date` | string | cycle date (key: series\|cycle) |
| `.windows.{key}.window` | string | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `.perEngineCycle.{series}.{cycle}.{engine}.{key}` | number | n: answers captured; viaSource: answers citing a qualifying source; viaText: answers reporting a readout result; either: union (the E1 count); unverified: answers citing an unreadable new page and no qualifying source |
| `.perEngineCycle.{series}.{cycle}.{engine}.{key}[]` | (empty) or string | readouts: readouts carried by the qualifying answers; readoutsUnverifiedAtCitation: readouts on a cited page not shown to be there when cited (D19) |
| `.latency[].series` | string | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `.latency[].engine` | string | engine: perplexity (Perplexity, sonar) \| google (Gemini, gemini-2.5-flash) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5) |
| `.latency[].readout` | string | readout label |
| `.latency[].presentedUtc` | string | presentation time, UTC |
| `.latency[].firstCycle` | null or string | first admitted cycle with a qualifying answer; null if none |
| `.latency[].days` | null or number | days from presentation to that cycle; null if not observed |
| `.latency[].censoredAtDay` | number or null | right-censoring day (not observed by day N); null when observed |
| `.crossArm[].answerSeries` | string | series of the answer |
| `.crossArm[].registeredSeries[]` | string | series the readout is registered for |
| `.crossArm[].engine` | string | engine: perplexity (Perplexity, sonar) \| google (Gemini, gemini-2.5-flash) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5) |
| `.crossArm[].readout` | string | readout |
| `.crossArm[].firstCycle` | string | cycle code: UAE-Cnn \| SG-Cnn \| SA-Cnn, a neutral identifier of one scan of one series' battery on four engines (the study's internal cycle identifiers are withheld, R5a; fields keep their frozen names, such as slug and cycleSlug) |
| `.crossArm[].days` | number | days from presentation |
| `.crossArm[].contentAtCitation` | string | verified \| unverified (D19) |
| `.revisedSources[].series` | string | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `.revisedSources[].url` | string | page URL |
| `.revisedSources[].class` | string | a-congress \| b-journal \| c-sponsor \| d-media \| not-congress-derived |
| `.revisedSources[].readouts[]` | string | readouts the page carries |
| `.revisedSources[].readoutsUnverifiedAtCitation[]` | string | readouts not shown to be on the page when cited |
| `.revisedSources[].contentAtCitation` | string | what the page can be shown to have said at citation |
| `.revisedSources[].adjudication` | string | the adjudication, verbatim |
| `.revisedSources[].revisions[].revisedAt` | string | UTC time of the revision |
| `.revisedSources[].revisions[].previousClass` | string | class before revision |
| `.revisedSources[].revisions[].previousReadouts[]` | string | readouts before revision |
| `.revisedSources[].revisions[].reason` | string | why it was revised |
| `.qualifyingAnswers.{sha256}[]` | string | readouts for which the answer qualifies (key: sha256 of the stored answer capture; links rows across files and to the held pull) |
| `.inspectionCoverage.note` | string | definition of coverage |
| `.inspectionCoverage.bySeriesEngineWindow.{key}.answers` | number | answers (key: series\|engine\|window) |
| `.inspectionCoverage.bySeriesEngineWindow.{key}.citingNewUrl` | number | answers citing at least one new URL |
| `.inspectionCoverage.bySeriesEngineWindow.{key}.citingUnreadable` | number | answers citing at least one unreadable new page |
| `.inspectionCoverage.bySeriesEngineWindow.{key}.newUrlCitations` | number | new-URL citations |
| `.inspectionCoverage.bySeriesEngineWindow.{key}.unreadableCitations` | number | citations of unreadable pages |
| `.inspectionCoverage.bySeriesEngineWindow.{key}.newUrlCitationsRead` | number | share of new-URL citations read, proportion |
| `.inspectionCoverage.unreadableTopDomains.{series}[][]` | string or number | [domain or token, unreadable citations] for the most frequent unreadable domains |

## `e1/easd-2026-e1.md`

- **Role:** e1-md. The E1 tables as printed by the tally run.
- **Endpoint:** E1
- **Source:** `docs/evidence/easd-2026-e1.md` (frozen); transformed (see REDACTION-LOG.md)
- **Structure:** per-series tables of qualifying answers per cycle and engine ("k/n (+u unverified)"), inspection coverage, most frequent unreadable domains (withheld ones as tokens), latency, cross-arm readouts, corrected classifications and the qualifying sources.

## `e4/easd-2026-e4-label-audit.json`

- **Role:** e4-label-audit. The seeded 150-figure audit of the scripted E4 attribution, with reason codes.
- **Endpoint:** E4 (quarantined, D17)
- **Source:** `docs/evidence/easd-2026-e4-label-audit.json` (frozen); transformed (see REDACTION-LOG.md)

| Field | Type | Meaning, units and allowed values |
|---|---|---|
| `.rule` | string | audit rule (D17) |
| `.reader` | string | who read |
| `.population` | number | label and published figures (1,082) |
| `.sample` | number | seeded sample (150; seed 20261010) |
| `.firstRead.wrong` | number | wrong on first read |
| `.firstRead.share` | number | proportion |
| `.firstRead.wilson95[]` | number | Wilson 95% interval |
| `.adjudicated.wrong` | number | wrong after adjudication |
| `.adjudicated.share` | number | proportion |
| `.adjudicated.wilson95[]` | number | Wilson 95% interval |
| `.verdict` | string | audit verdict |
| `.items[].id` | number | item |
| `.items[].sha256` | string | sha256 of the stored answer capture; links rows across files and to the held pull |
| `.items[].text` | string | figure as written (a marker for Gemini answers and beside a P61 sentence) |
| `.items[].value` | number or null | value; null beside a P61 sentence |
| `.items[].verdict` | string | correct \| wrong-attribution (first read) |
| `.items[].note` | string | note |
| `.items[].adjudicated` | string | correct \| wrong-attribution |
| `.items[].reason` | string | reason code or empty |
| `.items[].textSha256` | string | sha256 of the withheld Gemini figure string (present only where it is withheld, R1m) |
| `.items[].adjudicationNote` | string | adjudication note |
| `.reasonCodes.{key}` | string | reason code definition |

## `e4/easd-2026-e4-reference-values.json`

- **Role:** e4-reference-values. Label and published trial figures with their sources, written before any answer was coded.
- **Endpoint:** E4 reference values (fixed before coding)
- **Source:** `docs/evidence/easd-2026-e4-reference-values.json` (frozen); byte-identical to the source

| Field | Type | Meaning, units and allowed values |
|---|---|---|
| `.purpose` | string | purpose |
| `.protocol` | string | provenance string (script, protocol section) |
| `.fetched` | string | date fetched |
| `.sources.{key}.what` | string | source description |
| `.sources.{key}.url` | string | source URL |
| `.sources.{key}.locatedVia` | string | how it was located |
| `.measures.{key}` | string | measure definition |
| `.values[].id` | string | reference value id |
| `.values[].category` | string | label \| published \| congress |
| `.values[].source` | string | source key |
| `.values[].trial` | string | trial |
| `.values[].dose` | string | dose |
| `.values[].measure` | string | weight-pct \| weight-kg \| hba1c-pp |
| `.values[].value` | number | value (unit by measure) |
| `.values[].estimand` | string | estimand |
| `.values[].timepoint` | string | timepoint |
| `.values[].quote` | string | quotation from the source |
| `.values[].note` | string | note |

## `endpoints/ENDPOINT-SPECIFICATIONS.md`

- **Role:** endpoint-specifications. Every endpoint rule, reading and coding decision, copied from the public files.
- **Endpoint:** All
- **Source:** generated by the build; —
- **Structure:** the registration status of every endpoint, the registered rules (protocol §2–§6, §8), the readings I1–I5, the coding-scheme decisions S1–S11 and T1–T9, the E4 audit rule, and deviation D24 with the status of every post-registration robustness item, each copied from its public file.

## `figures/fig01-study-timeline.csv`

- **Role:** figure-data-fig01-study-timeline. The values plotted in the figure, written by figures.ts from the frozen record.
- **Endpoint:** Publication figure data
- **Source:** `docs/research/easd-2026/figures/fig01-study-timeline.csv` (committed, not frozen); byte-identical to the source

| Column | Meaning, units and allowed values |
|---|---|
| `kind` | row kind (fig01: window \| event \| readout \| cycle; fig04: cycle \| band \| pooled …) |
| `series` | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `id_or_label` | window label, event, readout or cycle code |
| `from_utc` | start, UTC |
| `to_utc` | end, UTC |
| `window` | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `engine_version` | instrument version 4 \| 5 |
| `admitted` | admitted under §2 (true \| false) |
| `status_or_reason` | registration status or exclusion reason |

## `figures/fig03-observation-flow.csv`

- **Role:** figure-data-fig03-observation-flow. The values plotted in the figure, written by figures.ts from the frozen record.
- **Endpoint:** Publication figure data
- **Source:** `docs/research/easd-2026/figures/fig03-observation-flow.csv` (committed, not frozen); byte-identical to the source

| Column | Meaning, units and allowed values |
|---|---|
| `series` | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `stage` | stage of the observation flow |
| `window` | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `count` | count |
| `unit` | unit of the count |
| `note` | note |

## `figures/fig04-uae-naming-trajectories.csv`

- **Role:** figure-data-fig04-uae-naming-trajectories. The values plotted in the figure, written by figures.ts from the frozen record.
- **Endpoint:** Publication figure data
- **Source:** `docs/research/easd-2026/figures/fig04-uae-naming-trajectories.csv` (committed, not frozen); byte-identical to the source

| Column | Meaning, units and allowed values |
|---|---|
| `series` | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `engine` | engine: perplexity (Perplexity, sonar) \| google (Gemini, gemini-2.5-flash) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5) \| all |
| `kind` | row kind (fig01: window \| event \| readout \| cycle; fig04: cycle \| band \| pooled …) |
| `cycle` | cycle code: UAE-Cnn \| SG-Cnn \| SA-Cnn, a neutral identifier of one scan of one series' battery on four engines (the study's internal cycle identifiers are withheld, R5a; fields keep their frozen names, such as slug and cycleSlug) |
| `scanned_at_or_from` | cycle completion time (UTC) or start of a pooled period |
| `window` | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `engine_version` | instrument version 4 \| 5 |
| `k` | named answers |
| `n` | governed answers |
| `rate` | k/n, proportion |
| `lo_or_band_low` | Wilson 95% lower bound (cycles) or band low (bands), proportion |
| `hi_or_band_high` | Wilson 95% upper bound or band high, proportion |
| `source` | file and field the value comes from |

## `figures/fig04b-saudi-naming-trajectories.csv`

- **Role:** figure-data-fig04b-saudi-naming-trajectories. The values plotted in the figure, written by figures.ts from the frozen record.
- **Endpoint:** Publication figure data
- **Source:** `docs/research/easd-2026/figures/fig04b-saudi-naming-trajectories.csv` (committed, not frozen); byte-identical to the source

| Column | Meaning, units and allowed values |
|---|---|
| `series` | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `engine` | engine: perplexity (Perplexity, sonar) \| google (Gemini, gemini-2.5-flash) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5) \| all |
| `kind` | row kind (fig01: window \| event \| readout \| cycle; fig04: cycle \| band \| pooled …) |
| `cycle` | cycle code: UAE-Cnn \| SG-Cnn \| SA-Cnn, a neutral identifier of one scan of one series' battery on four engines (the study's internal cycle identifiers are withheld, R5a; fields keep their frozen names, such as slug and cycleSlug) |
| `scanned_at_or_from` | cycle completion time (UTC) or start of a pooled period |
| `window` | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `engine_version` | instrument version 4 \| 5 |
| `k` | named answers |
| `n` | governed answers |
| `rate` | k/n, proportion |
| `lo_or_band_low` | Wilson 95% lower bound (cycles) or band low (bands), proportion |
| `hi_or_band_high` | Wilson 95% upper bound or band high, proportion |
| `source` | file and field the value comes from |

## `figures/fig05-e2-band-distance.csv`

- **Role:** figure-data-fig05-e2-band-distance. The values plotted in the figure, written by figures.ts from the frozen record.
- **Endpoint:** Publication figure data
- **Source:** `docs/research/easd-2026/figures/fig05-e2-band-distance.csv` (committed, not frozen); byte-identical to the source

| Column | Meaning, units and allowed values |
|---|---|
| `series` | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `engine` | engine: perplexity (Perplexity, sonar) \| google (Gemini, gemini-2.5-flash) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5) \| all |
| `registered` | registered arm (true \| false) |
| `after_k` | after-window named answers |
| `after_n` | after-window governed answers |
| `after_rate_exact` | after k/n unrounded |
| `bridged_low_as_decided` | bridged band low as decided |
| `bridged_high_as_decided` | bridged band high as decided |
| `bridged_low_exact_cycle` | exact-cycle bridged low (D19) |
| `bridged_high_exact_cycle` | exact-cycle bridged high (D19) |
| `nearer_edge` | nearer band edge |
| `distance_as_decided_pp` | signed distance, pp (negative inside) |
| `distance_exact_registered_band_pp` | signed distance, exact registered band, pp |
| `distance_exact_cycle_band_pp` | signed distance, exact-cycle band, pp |
| `cluster_ci95_low` | clustered 95% low |
| `cluster_ci95_high` | clustered 95% high |
| `cluster_ci95_contains_edge` | borderline flag |
| `movement_declared` | registered verdict |
| `movement_under_exact_cycle_band` | verdict under the exact-cycle band (robustness) |

## `figures/fig06-e1-inspection-coverage.csv`

- **Role:** figure-data-fig06-e1-inspection-coverage. The values plotted in the figure, written by figures.ts from the frozen record.
- **Endpoint:** Publication figure data
- **Source:** `docs/research/easd-2026/figures/fig06-e1-inspection-coverage.csv` (committed, not frozen); byte-identical to the source

| Column | Meaning, units and allowed values |
|---|---|
| `series` | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `engine` | engine: perplexity (Perplexity, sonar) \| google (Gemini, gemini-2.5-flash) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5) \| all |
| `window` | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `answers` | answers (count); in Figure 11, the row's answer group: citing the page \| not citing \| before \| after \| a decomposition or echo label; in addenda/PN1-response-structure.csv, the answers stored for that stream, engine and day |
| `answers_citing_new_url` | answers citing a new URL |
| `answers_citing_unreadable` | answers citing an unreadable page |
| `new_url_citations` | new-URL citations |
| `citations_read` | citations of readable pages |
| `citations_unreadable` | citations of unreadable pages |
| `share_read` | citations_read ÷ new_url_citations |

## `figures/fig07-e1-latency.csv`

- **Role:** figure-data-fig07-e1-latency. The values plotted in the figure, written by figures.ts from the frozen record.
- **Endpoint:** Publication figure data
- **Source:** `docs/research/easd-2026/figures/fig07-e1-latency.csv` (committed, not frozen); byte-identical to the source

| Column | Meaning, units and allowed values |
|---|---|
| `series` | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `readout` | readout |
| `presented_utc` | presentation time, UTC |
| `registration` | registered \| post-registration (Dn) |
| `engine` | engine: perplexity (Perplexity, sonar) \| google (Gemini, gemini-2.5-flash) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5) \| all |
| `status` | observed \| not observed \| unverified … |
| `days` | days from presentation |
| `censored_at_day` | right-censoring day |
| `first_cycle` | cycle code: UAE-Cnn \| SG-Cnn \| SA-Cnn, a neutral identifier of one scan of one series' battery on four engines (the study's internal cycle identifiers are withheld, R5a; fields keep their frozen names, such as slug and cycleSlug) |
| `note` | note |

## `figures/fig08-diatribe-source-to-claim.csv`

- **Role:** figure-data-fig08-diatribe-source-to-claim. The values plotted in the figure, written by figures.ts from the frozen record.
- **Endpoint:** Publication figure data
- **Source:** `docs/research/easd-2026/figures/fig08-diatribe-source-to-claim.csv` (committed, not frozen); byte-identical to the source

| Column | Meaning, units and allowed values |
|---|---|
| `element` | element of the source-to-claim chain |
| `field` | field of that element |
| `value` | value |
| `status` | observed \| not observed \| unverified … |
| `source` | file and field the value comes from |

## `figures/fig09-saudi-denominators.csv`

- **Role:** figure-data-fig09-saudi-denominators. The values plotted in the figure, written by figures.ts from the frozen record.
- **Endpoint:** Publication figure data
- **Source:** `docs/research/easd-2026/figures/fig09-saudi-denominators.csv` (committed, not frozen); byte-identical to the source

| Column | Meaning, units and allowed values |
|---|---|
| `path` | denominator path (A answers, B namings …) |
| `window` | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `stage` | stage of the observation flow |
| `value` | value |
| `unit_or_note` | unit or note |

## `figures/fig10-singapore-positioning.csv`

- **Role:** figure-data-fig10-singapore-positioning. The values plotted in the figure, written by figures.ts from the frozen record.
- **Endpoint:** Publication figure data
- **Source:** `docs/research/easd-2026/figures/fig10-singapore-positioning.csv` (committed, not frozen); byte-identical to the source

| Column | Meaning, units and allowed values |
|---|---|
| `cycle` | cycle code: UAE-Cnn \| SG-Cnn \| SA-Cnn, a neutral identifier of one scan of one series' battery on four engines (the study's internal cycle identifiers are withheld, R5a; fields keep their frozen names, such as slug and cycleSlug) |
| `date` | UTC date |
| `window` | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `engine_version` | instrument version 4 \| 5 |
| `avg_rank` | average rank of empagliflozin (Jardiance) in answers (descriptive, I4) |

## `figures/fig11-claude-cited-page.csv`

- **Role:** figure-data-fig11-claude-cited-page. The values plotted in the figure, written by figures.ts from the frozen record.
- **Endpoint:** Publication figure data
- **Source:** `docs/research/easd-2026/figures/fig11-claude-cited-page.csv` (committed, not frozen); byte-identical to the source

| Column | Meaning, units and allowed values |
|---|---|
| `cycle_or_group` | Figure 11 row: a cycle date, a pooled reference, a question group, the decomposition or the echo check |
| `window` | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `answers` | answers (count); in Figure 11, the row's answer group: citing the page \| not citing \| before \| after \| a decomposition or echo label; in addenda/PN1-response-structure.csv, the answers stored for that stream, engine and day |
| `k` | named answers |
| `n` | governed answers |
| `rate` | k/n, proportion |
| `wilson_low` | Wilson 95% lower bound, proportion |
| `wilson_high` | Wilson 95% upper bound, proportion |

## `figures/fig12-e2-band-readings.csv`

- **Role:** figure-data-fig12-e2-band-readings. The values plotted in the figure, written by figures.ts from the frozen record.
- **Endpoint:** Publication figure data
- **Source:** `docs/research/easd-2026/figures/fig12-e2-band-readings.csv` (committed, not frozen); byte-identical to the source

| Column | Meaning, units and allowed values |
|---|---|
| `analysis` | Figure 12 panel: band reading (item 2) or cycle-level change (item 3) |
| `series` | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `engine` | engine: perplexity (Perplexity, sonar) \| google (Gemini, gemini-2.5-flash) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5) \| all |
| `reading_or_variant` | reading of the band rule (i registered I1, ii shift and widen by half the CI width, iii shift and widen by the full width) or the change variant |
| `after_k_or_before_cycles` | after-window named answers, or the number of before cycles |
| `after_n_or_after_cycles` | after-window answers, or the number of after cycles |
| `after_rate` | after-window rate, proportion |
| `band_low` | band low under the reading, proportion |
| `band_high` | band high under the reading, proportion |
| `signed_distance_pp_or_change_pp` | signed distance to the nearer band edge (pp; negative inside), or the cycle-level change net of the offset (pp) |
| `movement` | movement under the reading (true \| false) |
| `flips` | whether the verdict differs from the registered reading |
| `ci95_low_pp` | 95% lower bound, pp |
| `ci95_high_pp` | 95% upper bound, pp |
| `z` | z statistic of the cycle-level change |

## `figures/fig13-source-class-mix.csv`

- **Role:** figure-data-fig13-source-class-mix. The values plotted in the figure, written by figures.ts from the frozen record.
- **Endpoint:** Publication figure data
- **Source:** `docs/research/easd-2026/figures/fig13-source-class-mix.csv` (committed, not frozen); byte-identical to the source

| Column | Meaning, units and allowed values |
|---|---|
| `view` | Figure 13 view: per-window mix or an onset comparison |
| `series` | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `engine` | engine: perplexity (Perplexity, sonar) \| google (Gemini, gemini-2.5-flash) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5) \| all |
| `window_or_pair` | window, or the pair of cycles compared |
| `class` | source class (regulator, journal, congress, sponsor, news-media, clinic, consumer-retail, government, academic, other, unclassified) |
| `links_in_class` | cited links in the class |
| `links_total` | cited links in the view |
| `share` | links_in_class ÷ links_total |
| `answers_or_earlier_share` | answers in the view, or the class share in the earlier cycle of a pair |

## `figures/FIGURES.md`

- **Role:** figures-captions. Caption, period, units, N and source of every figure.
- **Endpoint:** Publication figures
- **Source:** `docs/research/easd-2026/figures/FIGURES.md` (committed, not frozen); byte-identical to the source
- **Structure:** per figure: files, caption, observation period, units, N, source file and fields, and the key reading. The SVG figures it names are published with the paper. Figures 11–13 are post-registration robustness (D24).

## `freeze/easd-2026-freeze.json`

- **Role:** freeze. The freeze record: role, path, size and sha256 of every input and output, and of every study script.
- **Endpoint:** All
- **Source:** `docs/evidence/computed/easd-2026-freeze.json` (committed, not frozen); transformed (see REDACTION-LOG.md)

| Field | Type | Meaning, units and allowed values |
|---|---|---|
| `.generatedBy` | string | provenance string (script, protocol section) |
| `.frozenAt` | string | UTC time of the freeze |
| `.repositoryCommit` | string | commit at freeze |
| `.note` | string | what the record guarantees |
| `.files[].role` | string | role |
| `.files[].path` | string | path (the held pull's local path generalised to <OS temp>) |
| `.files[].bytes` | number | size |
| `.files[].sha256` | string | sha256 |
| `.files[].inRepository` | boolean | committed in the repository |
| `.files[].committedAsIs` | null or boolean | committed exactly as hashed (null for the pull) |
| `.files[].pull.pulledAt` | string | pull time |
| `.files[].pull.from` | string | first date |
| `.files[].pull.to` | string | last date |
| `.files[].pull.cycles` | number | cycles |
| `.files[].pull.bySeries.{series}` | number | cycles per series |
| `.scripts[].path` | string | script path |
| `.scripts[].sha256` | string | sha256 of the unsanitised script |
| `.scripts[].committedAsIs` | boolean | committed exactly as hashed |
| `.problems[]` | (empty) | problems found (empty) |

## `LICENCES.md`

- **Role:** licences. The asset-by-asset licence manifest (the release policy).
- **Endpoint:** All
- **Source:** generated by the build; —
- **Structure:** the licence summary under the release policy, the proprietary exclusions, a table giving every file its licence and the third-party material in it, and the MIT licence text for `code/`.

## `MANIFEST.json`

- **Role:** manifest. Path, role, bytes, sha256, source and transformations of every public file.
- **Endpoint:** All
- **Source:** generated by the build; —
- **Structure:** one entry per public file: path, role, group, endpoint, bytes, sha256, source path and sha256, whether the source matches the freeze record (with the frozen sha256 where a script was revised after the freeze), whether the copy is verbatim, the rules applied and the redaction counts; and the builder's own path and sha256.

## `protocol/easd-2026-evidence-to-answer-prereg.md`

- **Role:** protocol. The registered protocol (sections 1–10 byte-identical to registration commit 8ee24e19) and the append-only deviation log D0–D24.
- **Endpoint:** E1–E5, bridge, windows, admission, disclosure boundary; deviations D0–D24
- **Source:** `docs/design/easd-2026-evidence-to-answer-prereg.md` (frozen); transformed (see REDACTION-LOG.md)
- **Structure:** §1 question; §2 series, instrument and admission; §3 windows and the readout set; §4 endpoints E1–E5; §5 descriptive items; §6 bridge; §7 cadence; §8 reportable results; §9 deviations; §10 disclosure boundary; then the deviation log D0–D24, each dated with its reason and effect.

## `README.md`

- **Role:** readme. What the package is and how to use it.
- **Endpoint:** All
- **Source:** generated by the build; —
- **Structure:** what the package is, how it maps to the paper and supplement, three worked verification paths, the reproducibility table, the human-validation status, every restriction with its reason (question instrument, source identification, page P61, Gemini, full answers, operational details), the sanitised code, version identifiers, licence and third-party material, and correspondence.

## `redaction-domain-classification.csv`

- **Role:** domain-classification. The reviewable classification of every cited domain.
- **Endpoint:** All
- **Source:** generated by the build; —
- **Structure:** decision, type, entity (domain for named entities, token for withheld ones), citations in the E1 manifest, series, flagged and reason. Columns: decision (keep | withhold), type, entity, citations (answers citing pages of the entity in the manifest), series (series citing it, ; separated), flagged (true | false: unsure, operator not identifiable), reason (the rule or the reviewed decision under the source-identification policy).

## `REDACTION-LOG.md`

- **Role:** redaction-log. Every redaction rule, its counts and the domain classification.
- **Endpoint:** All
- **Source:** generated by the build; —
- **Structure:** every redaction rule with its count per file, the instrument-text items, the operational details, the cycle identifiers, the Gemini treatment, page P61, the source-identification policy and classification (named entities by name; withheld entities by token), the items flagged for human review, and the self-checks.

## `redaction-token-map.json`

- **Role:** token-map. Every withheld entity's token with its type, citations and series, and the unreadable-domain tables; tokens only, no names.
- **Endpoint:** All
- **Source:** generated by the build; —
- **Structure:** version and note; renamed: tokens renamed in this version by a type correction (from, to); entities[]: one object per withheld entity with token, type, local (true | false), citations (answers citing its pages in the E1 manifest), series[], flagged and reason; unreadableTopDomains: per series, [domain or token, unreadable citations] as in e1/easd-2026-e1.json. Tokens only, no names.

## `regulator/easd-2026-regulatory-status-2026-10-02.json`

- **Role:** regulator-status-2026-10-02. Regulator snapshot of 2 Oct used by the baseline coding and carried forward for the UAE.
- **Endpoint:** E3 status (UAE, Singapore), baseline coding
- **Source:** `docs/evidence/computed/easd-2026-regulatory-status-2026-10-02.json` (committed, not frozen); byte-identical to the source

| Field | Type | Meaning, units and allowed values |
|---|---|---|
| `.fetchedAt` | string | UTC time of the snapshot |
| `.readOnly` | boolean | true |
| `.rule` | string | status rule |
| `.entries[].agent` | string | agent |
| `.entries[].market` | string | UAE \| Singapore \| Saudi Arabia \| US \| EU |
| `.entries[].status` | string | approved \| not-approved \| not-established |
| `.entries[].approvedOn` | null or string | approval date; null if none |
| `.entries[].recordAsOf` | string | date the record establishes status to |
| `.entries[].regulator` | string | regulator and record |
| `.entries[].url` | string | record URL |
| `.entries[].qualifying` | string | qualifying text from the record |

## `regulator/easd-2026-regulatory-status-2026-10-11.json`

- **Role:** regulator-status. Regulator snapshot of 11 Oct used by the final coding.
- **Endpoint:** E3 status (UAE carried forward from 2 Oct; Singapore)
- **Source:** `docs/evidence/computed/easd-2026-regulatory-status-2026-10-11.json` (frozen); byte-identical to the source

| Field | Type | Meaning, units and allowed values |
|---|---|---|
| `.fetchedAt` | string | UTC time of the snapshot |
| `.readOnly` | boolean | true |
| `.rule` | string | status rule |
| `.previous.path` | string | snapshot carried forward |
| `.previous.fetchedAt` | string | its time |
| `.entries[].agent` | string | agent |
| `.entries[].market` | string | UAE \| Singapore \| Saudi Arabia \| US \| EU |
| `.entries[].status` | string | approved \| not-approved \| not-established |
| `.entries[].approvedOn` | null or string | approval date; null if none |
| `.entries[].recordAsOf` | string | date the record establishes status to |
| `.entries[].regulator` | string | regulator and record |
| `.entries[].url` | string | record URL |
| `.entries[].qualifying` | string | qualifying text from the record |

## `regulator/easd-2026-regulatory-status-saudi-t1d-2026-10-05.json`

- **Role:** regulator-status-saudi. SFDA, FDA and EMA records for the Saudi agent set, 5 Oct.
- **Endpoint:** Saudi E3 and E3-i status
- **Source:** `docs/evidence/computed/easd-2026-regulatory-status-saudi-t1d-2026-10-05.json` (frozen); byte-identical to the source

| Field | Type | Meaning, units and allowed values |
|---|---|---|
| `.fetchedAt` | string | UTC time of the snapshot |
| `.readOnly` | boolean | true |
| `.rule` | string | status rule |
| `.entries[].agent` | string | agent |
| `.entries[].set` | string | investigational \| indication (Saudi only) |
| `.entries[].market` | string | UAE \| Singapore \| Saudi Arabia \| US \| EU |
| `.entries[].status` | string | approved \| not-approved \| not-established |
| `.entries[].approvedOn` | null or string | approval date; null if none |
| `.entries[].recordAsOf` | string | date the record establishes status to |
| `.entries[].regulator` | string | regulator and record |
| `.entries[].url` | string | record URL |
| `.entries[].qualifying` | string | qualifying text from the record |

## `robustness/easd-2026-e1-corrections.csv`

- **Role:** e1-corrections. The E1 correction record: the answer the corrected classification treats as qualifying, in the per-answer table's public identifiers.
- **Endpoint:** All
- **Source:** generated by the build; —
- **Structure:** the E1 correction record: one row per answer the corrected classification treats as qualifying, keyed by the per-answer table's own public identifiers (series, cycleSlug, cycleDate, window, engine, promptId), with its governed and naming status, the frozen and corrected E1 flags and the reason. Use e1Qualifying alone for the frozen attribution calculation; additionally treat these rows as qualifying for the corrected one.

## `robustness/easd-2026-per-answer-outcomes.csv`

- **Role:** per-answer-outcomes. One row per captured answer in admitted cycles, without question or answer text; checked against every governed per-cycle count.
- **Endpoint:** POST-REGISTRATION ROBUSTNESS (D24) item 10: one row per captured answer (E2 naming, search, E1 flags)
- **Source:** `docs/evidence/computed/easd-2026-per-answer-outcomes.csv` (frozen); transformed (see REDACTION-LOG.md)
- **Header comment:** the first line starts with `#` and states the file's label and conventions; read the file with comment lines skipped.

| Column | Meaning, units and allowed values |
|---|---|
| `series` | series: uae-glp1 (Wegovy, UAE; registered) \| singapore-sglt2 (Jardiance, Singapore; registered) \| saudi-t1d (Tzield, Saudi Arabia; post-registration, D10) |
| `cycleSlug` | cycle code: UAE-Cnn \| SG-Cnn \| SA-Cnn, a neutral identifier of one scan of one series' battery on four engines (the study's internal cycle identifiers are withheld, R5a; fields keep their frozen names, such as slug and cycleSlug) |
| `cycleDate` | UTC date of the cycle |
| `window` | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `engine` | engine: perplexity (Perplexity, sonar) \| google (Gemini, gemini-2.5-flash) \| openai (ChatGPT, gpt-4o-mini) \| anthropic (Claude, claude-haiku-4-5) \| all |
| `promptId` | opaque question identifier (R2e): <uae\|sg\|sa>-<f\|q><NN>, the same map as the coding files |
| `focused` | focused question (asks about the measured medicine directly or head to head; true \| false) |
| `inGovernedPopulation` | counts in the E2 naming population (true \| false; focused questions are not) |
| `namedRegistered` | the production scorer's verdict: the answer names the brand or its molecule (the registered E2 measure; true \| false) |
| `namedBrandOnly` | the answer names the brand itself (no molecule credit; true \| false) |
| `searched` | the answer returned at least one cited link (true \| false) |
| `citationCount` | cited links in the answer (count) |
| `e1Qualifying` | the answer qualifies under E1 (true \| false; empty before the congress window, where E1 is not defined) |
| `citesUnverifiedNewPage` | the answer cites a new page that could not be read (true \| false; empty before the congress window) |
| `citesLocalPageP61` | the answer cites page P61, the local page behind Claude's UAE rise (true \| false; the page itself is withheld). Post-registration robustness (D24 item 12) |
| `echoesLocalPageSentence` | the answer shares a run of at least 40 normalised characters with one of page P61's two sentences (true \| false; the sentences are withheld). Post-registration robustness (D24 item 12) |

## `robustness/easd-2026-robustness.json`

- **Role:** robustness. The D24 post-registration robustness analyses, computed after the results were seen by a separate script; no registered verdict changes.
- **Endpoint:** POST-REGISTRATION ROBUSTNESS (D24): registered §5 analyses reported late, E2 band readings, cycle-level change, per-cycle view, corrected self-check, brand-only naming, search rate, attribution sensitivity, Saudi use context, E1 corrections, page P61
- **Source:** `docs/evidence/computed/easd-2026-robustness.json` (frozen); transformed (see REDACTION-LOG.md)

| Field | Type | Meaning, units and allowed values |
|---|---|---|
| `.label` | string | the D24 label carried by every output |
| `.deviation` | string | D24 |
| `.deterministic` | boolean | true: no clock reads |
| `.generatedBy` | string | provenance string (script, protocol section) |
| `.protocol` | string | provenance string (script, protocol section) |
| `.freeze.repositoryCommit` | string | the freeze in force when the robustness script ran (commit, time). It is the record superseded by the current freeze record; every frozen input it lists is byte-identical to the current record's (checked at build; README, Version identifiers) |
| `.freeze.frozenAt` | string | the freeze in force when the robustness script ran (commit, time). It is the record superseded by the current freeze record; every frozen input it lists is byte-identical to the current record's (checked at build; README, Version identifiers) |
| `.inputs.pull.path` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.pull.sha256` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.pull.frozen` | boolean | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.coding.path` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.coding.sha256` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.coding.frozen` | boolean | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.saudi.path` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.saudi.sha256` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.saudi.frozen` | boolean | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.bridge.path` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.bridge.sha256` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.bridge.frozen` | boolean | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.e1.path` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.e1.sha256` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.e1.frozen` | boolean | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.manifest.path` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.manifest.sha256` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.manifest.frozen` | boolean | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.ledger.path` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.ledger.sha256` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.ledger.frozen` | boolean | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.analysis.path` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.analysis.sha256` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.analysis.frozen` | boolean | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.manual.path` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.manual.sha256` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.manual.frozen` | boolean | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.note04.path` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.note04.sha256` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.note04.frozen` | boolean | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.note04.note` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.useContext.path` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.useContext.sha256` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.useContext.frozen` | boolean | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.useContext.note` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.geminiResolved.path` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.geminiResolved.sha256` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.geminiResolved.frozen` | boolean | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.inputs.geminiResolved.note` | string | each input: path (the pull and the Gemini redirect cache are local and identified by sha256 only), sha256, whether it is frozen, and a note |
| `.validation.governedRescoring.cycles` | number | the reproduction checks run first: every governed per-cycle count re-scored and matched; the E1 sets reproduced; registered cross-checks |
| `.validation.governedRescoring.cells` | number | the reproduction checks run first: every governed per-cycle count re-scored and matched; the E1 sets reproduced; registered cross-checks |
| `.validation.governedRescoring.mismatches` | number | the reproduction checks run first: every governed per-cycle count re-scored and matched; the E1 sets reproduced; registered cross-checks |
| `.validation.governedRescoring.note` | string | the reproduction checks run first: every governed per-cycle count re-scored and matched; the E1 sets reproduced; registered cross-checks |
| `.validation.e1Sets.reproduced` | boolean | the reproduction checks run first: every governed per-cycle count re-scored and matched; the E1 sets reproduced; registered cross-checks |
| `.validation.e1Sets.checked` | string | the reproduction checks run first: every governed per-cycle count re-scored and matched; the E1 sets reproduced; registered cross-checks |
| `.validation.e1Sets.mismatches` | number | the reproduction checks run first: every governed per-cycle count re-scored and matched; the E1 sets reproduced; registered cross-checks |
| `.validation.registeredCrossChecks` | string | the reproduction checks run first: every governed per-cycle count re-scored and matched; the E1 sets reproduced; registered cross-checks |
| `.item1RegisteredSection5ReportedLate.status` | string | registered §5 analyses, reported late under D24 |
| `.item1RegisteredSection5ReportedLate.a_note04ClinicPage.note` | string | §5 (a): per admitted UAE cycle and engine, answers, answers citing the Note 04 clinic page (never named), share, and those also naming the medicine; Gemini page-level matching only where redirects were resolved |
| `.item1RegisteredSection5ReportedLate.a_note04ClinicPage.perCycle[].slug` | string | §5 (a): per admitted UAE cycle and engine, answers, answers citing the Note 04 clinic page (never named), share, and those also naming the medicine; Gemini page-level matching only where redirects were resolved |
| `.item1RegisteredSection5ReportedLate.a_note04ClinicPage.perCycle[].date` | string | §5 (a): per admitted UAE cycle and engine, answers, answers citing the Note 04 clinic page (never named), share, and those also naming the medicine; Gemini page-level matching only where redirects were resolved |
| `.item1RegisteredSection5ReportedLate.a_note04ClinicPage.perCycle[].window` | string | §5 (a): per admitted UAE cycle and engine, answers, answers citing the Note 04 clinic page (never named), share, and those also naming the medicine; Gemini page-level matching only where redirects were resolved |
| `.item1RegisteredSection5ReportedLate.a_note04ClinicPage.perCycle[].{engine}.answers` | number | §5 (a): per admitted UAE cycle and engine, answers, answers citing the Note 04 clinic page (never named), share, and those also naming the medicine; Gemini page-level matching only where redirects were resolved |
| `.item1RegisteredSection5ReportedLate.a_note04ClinicPage.perCycle[].{engine}.citing` | number or null | §5 (a): per admitted UAE cycle and engine, answers, answers citing the Note 04 clinic page (never named), share, and those also naming the medicine; Gemini page-level matching only where redirects were resolved |
| `.item1RegisteredSection5ReportedLate.a_note04ClinicPage.perCycle[].{engine}.share` | number | §5 (a): per admitted UAE cycle and engine, answers, answers citing the Note 04 clinic page (never named), share, and those also naming the medicine; Gemini page-level matching only where redirects were resolved |
| `.item1RegisteredSection5ReportedLate.a_note04ClinicPage.perCycle[].{engine}.citingNamed` | number | §5 (a): per admitted UAE cycle and engine, answers, answers citing the Note 04 clinic page (never named), share, and those also naming the medicine; Gemini page-level matching only where redirects were resolved |
| `.item1RegisteredSection5ReportedLate.a_note04ClinicPage.perCycle[].{engine}.note` | string | §5 (a): per admitted UAE cycle and engine, answers, answers citing the Note 04 clinic page (never named), share, and those also naming the medicine; Gemini page-level matching only where redirects were resolved |
| `.item1RegisteredSection5ReportedLate.b_newLocalOrRegionalPages.rule` | string | §5 (b): new local or regional pages drawing ≥ 10% of one engine's answers in a cycle; each page has an opaque id (P-number), a type, region and counts; local businesses are typed, never named |
| `.item1RegisteredSection5ReportedLate.b_newLocalOrRegionalPages.rows[].series` | string | §5 (b): new local or regional pages drawing ≥ 10% of one engine's answers in a cycle; each page has an opaque id (P-number), a type, region and counts; local businesses are typed, never named |
| `.item1RegisteredSection5ReportedLate.b_newLocalOrRegionalPages.rows[].slug` | string | §5 (b): new local or regional pages drawing ≥ 10% of one engine's answers in a cycle; each page has an opaque id (P-number), a type, region and counts; local businesses are typed, never named |
| `.item1RegisteredSection5ReportedLate.b_newLocalOrRegionalPages.rows[].date` | string | §5 (b): new local or regional pages drawing ≥ 10% of one engine's answers in a cycle; each page has an opaque id (P-number), a type, region and counts; local businesses are typed, never named |
| `.item1RegisteredSection5ReportedLate.b_newLocalOrRegionalPages.rows[].window` | string | §5 (b): new local or regional pages drawing ≥ 10% of one engine's answers in a cycle; each page has an opaque id (P-number), a type, region and counts; local businesses are typed, never named |
| `.item1RegisteredSection5ReportedLate.b_newLocalOrRegionalPages.rows[].engine` | string | §5 (b): new local or regional pages drawing ≥ 10% of one engine's answers in a cycle; each page has an opaque id (P-number), a type, region and counts; local businesses are typed, never named |
| `.item1RegisteredSection5ReportedLate.b_newLocalOrRegionalPages.rows[].page` | string | §5 (b): new local or regional pages drawing ≥ 10% of one engine's answers in a cycle; each page has an opaque id (P-number), a type, region and counts; local businesses are typed, never named |
| `.item1RegisteredSection5ReportedLate.b_newLocalOrRegionalPages.rows[].type` | string | §5 (b): new local or regional pages drawing ≥ 10% of one engine's answers in a cycle; each page has an opaque id (P-number), a type, region and counts; local businesses are typed, never named |
| `.item1RegisteredSection5ReportedLate.b_newLocalOrRegionalPages.rows[].answersCiting` | number | §5 (b): new local or regional pages drawing ≥ 10% of one engine's answers in a cycle; each page has an opaque id (P-number), a type, region and counts; local businesses are typed, never named |
| `.item1RegisteredSection5ReportedLate.b_newLocalOrRegionalPages.rows[].answers` | number | §5 (b): new local or regional pages drawing ≥ 10% of one engine's answers in a cycle; each page has an opaque id (P-number), a type, region and counts; local businesses are typed, never named |
| `.item1RegisteredSection5ReportedLate.b_newLocalOrRegionalPages.rows[].share` | number | §5 (b): new local or regional pages drawing ≥ 10% of one engine's answers in a cycle; each page has an opaque id (P-number), a type, region and counts; local businesses are typed, never named |
| `.item1RegisteredSection5ReportedLate.b_newLocalOrRegionalPages.rows[].namedAmongCiting` | number | §5 (b): new local or regional pages drawing ≥ 10% of one engine's answers in a cycle; each page has an opaque id (P-number), a type, region and counts; local businesses are typed, never named |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.unit` | string | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.distinctHosts` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.links` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.byClass.regulator.hosts` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.byClass.regulator.links` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.byClass.journal.hosts` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.byClass.journal.links` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.byClass.congress.hosts` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.byClass.congress.links` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.byClass.sponsor.hosts` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.byClass.sponsor.links` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.byClass.news-media.hosts` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.byClass.news-media.links` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.byClass.clinic.hosts` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.byClass.clinic.links` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.byClass.consumer-retail.hosts` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.byClass.consumer-retail.links` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.byClass.other.hosts` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.byClass.other.links` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.byClass.unclassified.hosts` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.byClass.unclassified.links` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.otherBySubclass.academic` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.otherBySubclass.directory-or-platform` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.otherBySubclass.government-non-health` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.otherBySubclass.insurer` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.otherBySubclass.market-research` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.otherBySubclass.professional-or-patient-organisation` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.otherBySubclass.reference` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.otherBySubclass.social-or-platform` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.unclassifiedLinkShare` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.handAssignedHashes` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.namedPublicHosts` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.geminiLinks` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.classifierCoverage.geminiLinksWithoutHostTitle` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].series` | string | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].window` | string | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].engine` | string | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].answers` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].citingAnswers` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].links` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].counts.regulator` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].counts.journal` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].counts.congress` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].counts.sponsor` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].counts.news-media` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].counts.clinic` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].counts.consumer-retail` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].counts.other` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].counts.unclassified` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].shares.regulator` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].shares.journal` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].shares.congress` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].shares.sponsor` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].shares.news-media` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].shares.clinic` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].shares.consumer-retail` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].shares.other` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].shares.unclassified` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].answersCiting.regulator` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].answersCiting.journal` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].answersCiting.congress` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].answersCiting.sponsor` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].answersCiting.news-media` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].answersCiting.clinic` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].answersCiting.consumer-retail` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].answersCiting.other` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].answersCiting.unclassified` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].namedAmongAnswersCiting.regulator` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].namedAmongAnswersCiting.journal` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].namedAmongAnswersCiting.congress` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].namedAmongAnswersCiting.sponsor` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].namedAmongAnswersCiting.news-media` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].namedAmongAnswersCiting.clinic` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].namedAmongAnswersCiting.consumer-retail` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].namedAmongAnswersCiting.other` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.rows[].namedAmongAnswersCiting.unclassified` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].label` | string | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].engine` | string | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.slug` | string | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.date` | string | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.governedNaming.k` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.governedNaming.n` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.searched` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.answers` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.citingAnswers` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.links` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.counts.regulator` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.counts.journal` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.counts.congress` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.counts.sponsor` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.counts.news-media` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.counts.clinic` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.counts.consumer-retail` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.counts.other` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.counts.unclassified` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.shares.regulator` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.shares.journal` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.shares.congress` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.shares.sponsor` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.shares.news-media` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.shares.clinic` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.shares.consumer-retail` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.shares.other` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.shares.unclassified` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.answersCiting.regulator` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.answersCiting.journal` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.answersCiting.congress` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.answersCiting.sponsor` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.answersCiting.news-media` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.answersCiting.clinic` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.answersCiting.consumer-retail` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.answersCiting.other` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.answersCiting.unclassified` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.namedAmongAnswersCiting.regulator` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.namedAmongAnswersCiting.journal` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.namedAmongAnswersCiting.congress` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.namedAmongAnswersCiting.sponsor` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.namedAmongAnswersCiting.news-media` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.namedAmongAnswersCiting.clinic` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.namedAmongAnswersCiting.consumer-retail` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.namedAmongAnswersCiting.other` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].from.namedAmongAnswersCiting.unclassified` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.slug` | string | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.date` | string | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.governedNaming.k` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.governedNaming.n` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.searched` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.answers` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.citingAnswers` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.links` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.counts.regulator` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.counts.journal` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.counts.congress` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.counts.sponsor` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.counts.news-media` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.counts.clinic` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.counts.consumer-retail` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.counts.other` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.counts.unclassified` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.shares.regulator` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.shares.journal` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.shares.congress` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.shares.sponsor` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.shares.news-media` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.shares.clinic` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.shares.consumer-retail` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.shares.other` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.shares.unclassified` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.answersCiting.regulator` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.answersCiting.journal` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.answersCiting.congress` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.answersCiting.sponsor` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.answersCiting.news-media` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.answersCiting.clinic` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.answersCiting.consumer-retail` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.answersCiting.other` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.answersCiting.unclassified` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.namedAmongAnswersCiting.regulator` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.namedAmongAnswersCiting.journal` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.namedAmongAnswersCiting.congress` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.namedAmongAnswersCiting.sponsor` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.namedAmongAnswersCiting.news-media` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.namedAmongAnswersCiting.clinic` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.namedAmongAnswersCiting.consumer-retail` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.namedAmongAnswersCiting.other` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].to.namedAmongAnswersCiting.unclassified` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].shareChangePp.regulator` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].shareChangePp.journal` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].shareChangePp.congress` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].shareChangePp.sponsor` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].shareChangePp.news-media` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].shareChangePp.clinic` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].shareChangePp.consumer-retail` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].shareChangePp.other` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].shareChangePp.unclassified` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].distinctDomains.from` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].distinctDomains.to` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].distinctDomains.newInLater` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].distinctDomains.absentFromLater` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].topNewDomains[].domain` | string | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].topNewDomains[].class` | string | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].topNewDomains[].answersCiting` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].topNewDomains[].namedAmongCiting` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].topDroppedDomains[].domain` | string | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].topDroppedDomains[].class` | string | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item1RegisteredSection5ReportedLate.c_sourceClassMix.comparisons[].topDroppedDomains[].answersCitingInEarlier` | number | §5 (c): cited links by source class per engine, window and series (counts and shares), with the two onset comparisons (Claude UAE 27 vs 29 Sep; Perplexity UAE 6 vs 7 Oct); classes by host rule, public entities named, local businesses typed |
| `.item2BandReadings.readings.i-registered-I1` | string | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.readings.ii-shift-widen-half-width` | string | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.readings.iii-shift-widen-full-width` | string | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.readings.reference-shift-only` | string | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].series` | string | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].registered` | boolean | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].engine` | string | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].{window}.k` | number | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].{window}.n` | number | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].{window}.rate` | number | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].v4Band[]` | number | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].offset.pairedQuestions` | number | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].offset.offset` | number | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].offset.ci95[]` | number | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].readings.i-registered-I1.band[]` | number | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].readings.i-registered-I1.movement` | boolean | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].readings.i-registered-I1.signedDistancePp` | number | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].readings.ii-shift-widen-half-width.band[]` | number | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].readings.ii-shift-widen-half-width.movement` | boolean | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].readings.ii-shift-widen-half-width.signedDistancePp` | number | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].readings.iii-shift-widen-full-width.band[]` | number | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].readings.iii-shift-widen-full-width.movement` | boolean | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].readings.iii-shift-widen-full-width.signedDistancePp` | number | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].readings.reference-shift-only.band[]` | number | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].readings.reference-shift-only.movement` | boolean | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].readings.reference-shift-only.signedDistancePp` | number | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].registeredVerdict` | string | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item2BandReadings.rows[].flipsUnderReadings[]` | string | E2 under three readings of "widened by the offset's 95% interval": the band, verdict and signed distance (pp) per series and engine for each reading, and which verdicts flip |
| `.item3CycleLevelChange.estimand` | string | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.seMethod` | string | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].series` | string | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].registered` | boolean | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].engine` | string | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].beforeCycles` | number | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].beforeReference` | string | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].beforeMeanRate` | number | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].beforeSdPp` | number | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].afterCycles` | number | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].afterMeanRate` | number | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].afterSdPp` | number | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].offsetPp` | number | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].offsetCi95Pp[]` | number | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].changeNetOfOffsetPp` | number | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].seFromCyclesPp` | number | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].seFromOffsetPp` | number | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].seFromOffsetConservativePp` | number | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].primary.sePp` | number | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].primary.z` | number | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].primary.ci95Pp[]` | number | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].primary.pTwoSided` | number | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].conservativeOffsetSe.sePp` | number | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].conservativeOffsetSe.z` | number | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].conservativeOffsetSe.ci95Pp[]` | number | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item3CycleLevelChange.rows[].conservativeOffsetSe.pTwoSided` | number | cycle-level change: mean after-cycle rate minus mean v4 reference-cycle rate minus the bridge offset, with SE, 95% CI (pp) and z; cycles are the unit |
| `.item4PerCycleAgainstBridgedBand.band` | string | each during- and after-window cycle (and the 27 Sep v5 cycle) against the bridged band: k, n, rate, Wilson 95% interval (too narrow: answers to one question correlate), inside or outside |
| `.item4PerCycleAgainstBridgedBand.interval` | string | each during- and after-window cycle (and the 27 Sep v5 cycle) against the bridged band: k, n, rate, Wilson 95% interval (too narrow: answers to one question correlate), inside or outside |
| `.item4PerCycleAgainstBridgedBand.rows[].series` | string | each during- and after-window cycle (and the 27 Sep v5 cycle) against the bridged band: k, n, rate, Wilson 95% interval (too narrow: answers to one question correlate), inside or outside |
| `.item4PerCycleAgainstBridgedBand.rows[].registered` | boolean | each during- and after-window cycle (and the 27 Sep v5 cycle) against the bridged band: k, n, rate, Wilson 95% interval (too narrow: answers to one question correlate), inside or outside |
| `.item4PerCycleAgainstBridgedBand.rows[].engine` | string | each during- and after-window cycle (and the 27 Sep v5 cycle) against the bridged band: k, n, rate, Wilson 95% interval (too narrow: answers to one question correlate), inside or outside |
| `.item4PerCycleAgainstBridgedBand.rows[].bridgedBand[]` | number | each during- and after-window cycle (and the 27 Sep v5 cycle) against the bridged band: k, n, rate, Wilson 95% interval (too narrow: answers to one question correlate), inside or outside |
| `.item4PerCycleAgainstBridgedBand.rows[].cycles[].slug` | string | each during- and after-window cycle (and the 27 Sep v5 cycle) against the bridged band: k, n, rate, Wilson 95% interval (too narrow: answers to one question correlate), inside or outside |
| `.item4PerCycleAgainstBridgedBand.rows[].cycles[].date` | string | each during- and after-window cycle (and the 27 Sep v5 cycle) against the bridged band: k, n, rate, Wilson 95% interval (too narrow: answers to one question correlate), inside or outside |
| `.item4PerCycleAgainstBridgedBand.rows[].cycles[].window` | string | each during- and after-window cycle (and the 27 Sep v5 cycle) against the bridged band: k, n, rate, Wilson 95% interval (too narrow: answers to one question correlate), inside or outside |
| `.item4PerCycleAgainstBridgedBand.rows[].cycles[].k` | number | each during- and after-window cycle (and the 27 Sep v5 cycle) against the bridged band: k, n, rate, Wilson 95% interval (too narrow: answers to one question correlate), inside or outside |
| `.item4PerCycleAgainstBridgedBand.rows[].cycles[].n` | number | each during- and after-window cycle (and the 27 Sep v5 cycle) against the bridged band: k, n, rate, Wilson 95% interval (too narrow: answers to one question correlate), inside or outside |
| `.item4PerCycleAgainstBridgedBand.rows[].cycles[].rate` | number | each during- and after-window cycle (and the 27 Sep v5 cycle) against the bridged band: k, n, rate, Wilson 95% interval (too narrow: answers to one question correlate), inside or outside |
| `.item4PerCycleAgainstBridgedBand.rows[].cycles[].wilson95[]` | number | each during- and after-window cycle (and the 27 Sep v5 cycle) against the bridged band: k, n, rate, Wilson 95% interval (too narrow: answers to one question correlate), inside or outside |
| `.item4PerCycleAgainstBridgedBand.rows[].cycles[].position` | string | each during- and after-window cycle (and the 27 Sep v5 cycle) against the bridged band: k, n, rate, Wilson 95% interval (too narrow: answers to one question correlate), inside or outside |
| `.item4PerCycleAgainstBridgedBand.rows[].cycles[].signedDistancePp` | number | each during- and after-window cycle (and the 27 Sep v5 cycle) against the bridged band: k, n, rate, Wilson 95% interval (too narrow: answers to one question correlate), inside or outside |
| `.item4PerCycleAgainstBridgedBand.rows[].cycles[].wilsonClearOfBand` | boolean | each during- and after-window cycle (and the 27 Sep v5 cycle) against the bridged band: k, n, rate, Wilson 95% interval (too narrow: answers to one question correlate), inside or outside |
| `.item4PerCycleAgainstBridgedBand.rows[].afterOutside` | number | each during- and after-window cycle (and the 27 Sep v5 cycle) against the bridged band: k, n, rate, Wilson 95% interval (too narrow: answers to one question correlate), inside or outside |
| `.item4PerCycleAgainstBridgedBand.rows[].afterAbove` | number | each during- and after-window cycle (and the 27 Sep v5 cycle) against the bridged band: k, n, rate, Wilson 95% interval (too narrow: answers to one question correlate), inside or outside |
| `.item4PerCycleAgainstBridgedBand.rows[].afterBelow` | number | each during- and after-window cycle (and the 27 Sep v5 cycle) against the bridged band: k, n, rate, Wilson 95% interval (too narrow: answers to one question correlate), inside or outside |
| `.item4PerCycleAgainstBridgedBand.rows[].afterCycles` | number | each during- and after-window cycle (and the 27 Sep v5 cycle) against the bridged band: k, n, rate, Wilson 95% interval (too narrow: answers to one question correlate), inside or outside |
| `.item5CorrectedBridgeSelfCheck.method` | string | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.d2.2` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.d2.3` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.d2.4` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.d2.5` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.d2.6` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.d2.7` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.d2.8` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.d2.9` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.d2.10` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.d2.11` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.d2.12` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.d2.13` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.d2.14` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.d2.15` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].series` | string | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].registered` | boolean | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].engine` | string | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].v4CyclesDefiningBand` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].bandWidthExactPp` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].bandWidthRegisteredPp` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].v4Runs[]` | string | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].v5Runs[]` | string | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].v4SpreadPp` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].v5SpreadPp` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].registeredRatio` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].expectedRatioNoDayEffect` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].d2AdjustedRatio.v4Runs` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].d2AdjustedRatio.v5Runs` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].d2AdjustedRatio.meanOfV4AndV5Ranges` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].sdBetweenCyclesPp` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].sdWithinDayV4Pp` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].sdWithinDayV5Pp` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].sdWithinDayPooledPp` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].fTestV4Only.F` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].fTestV4Only.df[]` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].fTestV4Only.pOneSided` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].fTestV4Only.note` | string | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].fTestPooledV4V5.F` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].fTestPooledV4V5.df[]` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].fTestPooledV4V5.pOneSided` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].fTestPooledV4V5.note` | string | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].binomialSdAtBandPeriodPp` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].binomialSdAtBridgeDayPp` | number | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].excessBetweenDayVariation.materialOnV4Runs` | boolean | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].excessBetweenDayVariation.materialOnV4AndV5Runs` | boolean | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.rows[].excessBetweenDayVariation.fTestP05` | boolean | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.reviewersExpectation[].series` | string | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.reviewersExpectation[].engine` | string | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.reviewersExpectation[].reviewersExpectExcess` | boolean | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.reviewersExpectation[].materialOnV4Runs` | boolean | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.reviewersExpectation[].materialOnV4AndV5Runs` | boolean | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.reviewersExpectation[].fTestP05` | boolean | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.reviewersExpectation[].matchesOnV4Runs` | boolean | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item5CorrectedBridgeSelfCheck.reviewersExpectation[].matchesOnV4AndV5Runs` | boolean | the bridge self-check corrected for sample size: d2 constants, d2-adjusted range ratios, between-day and within-day SDs, v5 spreads, per series and engine |
| `.item6BrandOnlyNaming.definitions.{series}.brand` | string | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.definitions.{series}.molecule` | string | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.definitions.{series}.registeredNames[]` | string | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.definitions.{series}.brandOnlyNames[]` | string | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.definitions.{series}.moleculeCredit[]` | string | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.rule` | string | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.rows[].series` | string | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.rows[].window` | string | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.rows[].engine` | string | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.rows[].n` | number | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.rows[].registeredK` | number | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.rows[].registeredRate` | number | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.rows[].brandOnlyK` | number | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.rows[].brandOnlyRate` | number | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.rows[].moleculeOnlyK` | number | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.rows[].brandOnlyNotRegistered` | number | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.rows[].canonicalFlagDisagreements` | number | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.perCycleUae[].slug` | string | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.perCycleUae[].date` | string | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.perCycleUae[].window` | string | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.perCycleUae[].byEngine.{engine}.n` | number | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.perCycleUae[].byEngine.{engine}.registeredK` | number | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item6BrandOnlyNaming.perCycleUae[].byEngine.{engine}.brandOnlyK` | number | brand-only naming (no molecule credit) beside the registered brand-or-molecule measure: definitions, per series, engine and window k/n, and per UAE cycle |
| `.item7SearchRate.definition` | string | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.perApi[].engine` | string | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.perApi[].endpoint` | string | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.perApi[].citation` | string | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.perApi[].searchResults` | string | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.rows[].series` | string | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.rows[].window` | string | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.rows[].engine` | string | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.rows[].answers` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.rows[].searched` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.rows[].searchRate` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.rows[].withSearchResults` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.rows[].definitionsCoincide` | boolean | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.rows[].meanCitationsPerSearchingAnswer` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.rows[].meanDistinctDomainsPerSearchingAnswer` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.rows[].governedAnswers` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.rows[].governedSearched` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.rows[].governedSearchRate` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.rows[].namingAmongSearching.k` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.rows[].namingAmongSearching.n` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.rows[].namingAmongNotSearching` | null | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.rows[].namingAmongNotSearching.k` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.rows[].namingAmongNotSearching.n` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.perCycleUae[].slug` | string | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.perCycleUae[].date` | string | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.perCycleUae[].window` | string | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.perCycleUae[].byEngine.{engine}.answers` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.perCycleUae[].byEngine.{engine}.searched` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.perCycleUae[].byEngine.{engine}.governed` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.perCycleUae[].byEngine.{engine}.governedSearched` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.perCycleUae[].byEngine.{engine}.namedAmongGovernedSearched` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.perCycleUae[].byEngine.{engine}.namedAmongGovernedNotSearched` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.attributionAmongSearching[].series` | string | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.attributionAmongSearching[].engine` | string | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.attributionAmongSearching[].allGoverned.qualifying` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.attributionAmongSearching[].allGoverned.qualifyingNamed` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.attributionAmongSearching[].allGoverned.others` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.attributionAmongSearching[].allGoverned.othersNamed` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.attributionAmongSearching[].allGoverned.fisherP` | number or null | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.attributionAmongSearching[].allGoverned.minAttainableP` | number or null | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.attributionAmongSearching[].allGoverned.qualifyingNeededForP05` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.attributionAmongSearching[].searchingOnly.qualifying` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.attributionAmongSearching[].searchingOnly.qualifyingNamed` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.attributionAmongSearching[].searchingOnly.others` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.attributionAmongSearching[].searchingOnly.othersNamed` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.attributionAmongSearching[].searchingOnly.fisherP` | number or null | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.attributionAmongSearching[].searchingOnly.minAttainableP` | number or null | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.attributionAmongSearching[].searchingOnly.qualifyingNeededForP05` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.attributionAmongSearching[].notSearching.answers` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item7SearchRate.attributionAmongSearching[].notSearching.named` | number | search rate (answers returning at least one cited link) by engine, window and series; per API; per UAE cycle; and E2 attribution among answers that searched |
| `.item8AttributionSensitivity.rule` | string | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].series` | string | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].engine` | string | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].registeredVerdict` | string | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].registered.qualifying` | number | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].registered.qualifyingNamed` | number | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].registered.others` | number | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].registered.othersNamed` | number | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].registered.fisherP` | number or null | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].registered.minAttainableP` | number or null | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].registered.qualifyingNeededForP05` | number | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].unverifiedOthersExcluded` | number | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].unverifiedOthersExcludedNamed` | number | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].unverifiedExcluded.qualifying` | number | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].unverifiedExcluded.qualifyingNamed` | number | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].unverifiedExcluded.others` | number | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].unverifiedExcluded.othersNamed` | number | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].unverifiedExcluded.fisherP` | number or null | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].unverifiedExcluded.minAttainableP` | number or null | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].unverifiedExcluded.qualifyingNeededForP05` | number or null | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].focusedQuestionsIncluded.qualifying` | number | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].focusedQuestionsIncluded.qualifyingNamed` | number | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].focusedQuestionsIncluded.others` | number | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].focusedQuestionsIncluded.othersNamed` | number | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].focusedQuestionsIncluded.fisherP` | number or null | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].focusedQuestionsIncluded.minAttainableP` | number or null | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.rows[].focusedQuestionsIncluded.qualifyingNeededForP05` | number | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.e1QualifyingByQuestionClass[].series` | string | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.e1QualifyingByQuestionClass[].window` | string | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.e1QualifyingByQuestionClass[].engine` | string | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.e1QualifyingByQuestionClass[].qualifying` | number | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.e1QualifyingByQuestionClass[].focused` | number | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.e1QualifyingByQuestionClass[].brandNeutral` | number | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.e1QualifyingByQuestionClass[].answers` | number | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item8AttributionSensitivity.e1QualifyingByQuestionClass[].focusedAnswers` | number | attribution with answers citing an unverified page removed from the comparison group; the minimum attainable Fisher p; E1 qualifying answers by question class |
| `.item9SaudiE3iUseContext.status` | string | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.codesFile` | string | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rules[]` | string | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.intervals` | string | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].window` | string | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].byContext.glycaemic` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].byContext.weight-management` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].byContext.unspecified` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].allT1dUse.namings` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].allT1dUse.distinctAnswers` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].allT1dUse.noStatusStatement` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].allT1dUse.share` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].allT1dUse.clusterCI95[]` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].allT1dUse.questions` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].glycaemic.namings` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].glycaemic.distinctAnswers` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].glycaemic.noStatusStatement` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].glycaemic.share` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].glycaemic.clusterCI95[]` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].glycaemic.questions` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].glycaemicNoWeightFraming.namings` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].glycaemicNoWeightFraming.distinctAnswers` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].glycaemicNoWeightFraming.noStatusStatement` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].glycaemicNoWeightFraming.share` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].glycaemicNoWeightFraming.clusterCI95[]` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].glycaemicNoWeightFraming.questions` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].glycaemicNotOverweightPopulation.namings` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].glycaemicNotOverweightPopulation.distinctAnswers` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].glycaemicNotOverweightPopulation.noStatusStatement` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].glycaemicNotOverweightPopulation.share` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].glycaemicNotOverweightPopulation.clusterCI95[]` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].glycaemicNotOverweightPopulation.questions` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].unspecified.namings` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].unspecified.distinctAnswers` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].unspecified.noStatusStatement` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].unspecified.share` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].unspecified.clusterCI95[]` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item9SaudiE3iUseContext.rows[].unspecified.questions` | number | Saudi E3-i restricted by use context (model-coded, same family): rules, codes file, per window and context the no-status share with question-clustered intervals |
| `.item10PerAnswerTable.file` | string | the per-answer outcome table: file, rows, cycles and governed cells checked, mismatches (0) and columns |
| `.item10PerAnswerTable.rows` | number | the per-answer outcome table: file, rows, cycles and governed cells checked, mismatches (0) and columns |
| `.item10PerAnswerTable.cyclesChecked` | number | the per-answer outcome table: file, rows, cycles and governed cells checked, mismatches (0) and columns |
| `.item10PerAnswerTable.governedCellsChecked` | number | the per-answer outcome table: file, rows, cycles and governed cells checked, mismatches (0) and columns |
| `.item10PerAnswerTable.mismatches` | number | the per-answer outcome table: file, rows, cycles and governed cells checked, mismatches (0) and columns |
| `.item10PerAnswerTable.columns[]` | string | the per-answer outcome table: file, rows, cycles and governed cells checked, mismatches (0) and columns |
| `.item10PerAnswerTable.item12Gates.reproduced` | boolean | the per-answer outcome table: file, rows, cycles and governed cells checked, mismatches (0) and columns |
| `.item10PerAnswerTable.item12Gates.windowsChecked` | number | the per-answer outcome table: file, rows, cycles and governed cells checked, mismatches (0) and columns |
| `.item10PerAnswerTable.item12Gates.cyclesChecked` | number | the per-answer outcome table: file, rows, cycles and governed cells checked, mismatches (0) and columns |
| `.item10PerAnswerTable.item12Gates.afterWindow.citingNamed` | string | the per-answer outcome table: file, rows, cycles and governed cells checked, mismatches (0) and columns |
| `.item10PerAnswerTable.item12Gates.afterWindow.notCitingNamed` | string | the per-answer outcome table: file, rows, cycles and governed cells checked, mismatches (0) and columns |
| `.item10PerAnswerTable.item12Gates.cycle29Sep.citingNamed` | string | the per-answer outcome table: file, rows, cycles and governed cells checked, mismatches (0) and columns |
| `.item10PerAnswerTable.item12Gates.cycle29Sep.notCitingNamed` | string | the per-answer outcome table: file, rows, cycles and governed cells checked, mismatches (0) and columns |
| `.item10PerAnswerTable.item12Gates.echoFrom29Sep.citingEchoing` | string | the per-answer outcome table: file, rows, cycles and governed cells checked, mismatches (0) and columns |
| `.item10PerAnswerTable.item12Gates.echoFrom29Sep.notCitingEchoing` | string | the per-answer outcome table: file, rows, cycles and governed cells checked, mismatches (0) and columns |
| `.item11E1Corrections.a_latencyIntervals.note` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.a_latencyIntervals.rows[].series` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.a_latencyIntervals.rows[].engine` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.a_latencyIntervals.rows[].readout` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.a_latencyIntervals.rows[].presentedUtc` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.a_latencyIntervals.rows[].firstCycle` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.a_latencyIntervals.rows[].firstCycleCompletedUtc` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.a_latencyIntervals.rows[].observedDay` | number | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.a_latencyIntervals.rows[].previousAdmittedCycle` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.a_latencyIntervals.rows[].previousAdmittedCycleCompletedUtc` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.a_latencyIntervals.rows[].previousCycleDay` | number | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.a_latencyIntervals.rows[].interval.lowerExclusiveDay` | number | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.a_latencyIntervals.rows[].interval.upperInclusiveDay` | number | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.a_latencyIntervals.rows[].previousCycleBeforePresentation` | boolean | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.a_latencyIntervals.rows[].excludedCyclesBetween[]` | (empty) | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.a_latencyIntervals.rows[].excludedCyclesBetween[].slug` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.a_latencyIntervals.rows[].excludedCyclesBetween[].date` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.a_latencyIntervals.rows[].excludedCyclesBetween[].reasons[]` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.b_textRouteMiss.note` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.b_textRouteMiss.rows[].answer` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.b_textRouteMiss.rows[].slug` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.b_textRouteMiss.rows[].date` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.b_textRouteMiss.rows[].window` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.b_textRouteMiss.rows[].engine` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.b_textRouteMiss.rows[].figure` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.b_textRouteMiss.rows[].figureSha256` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.b_textRouteMiss.rows[].reference[]` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.b_textRouteMiss.rows[].d4HandCheck[]` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.b_textRouteMiss.rows[].confirmedInPull` | boolean | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.b_textRouteMiss.rows[].answerNamesOctane` | boolean | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.b_textRouteMiss.rows[].e1Qualifying[]` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.b_textRouteMiss.rows[].e1QualifyingRoute` | string or null | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.b_textRouteMiss.rows[].inTextLedger` | boolean | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.b_textRouteMiss.rows[].textLedgerQualifies` | null | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.b_textRouteMiss.rows[].e1Qualifying` | null | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.b_textRouteMiss.genuine` | number | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.b_textRouteMiss.notFlaggedByTextRoute` | number | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.c_singaporeE3.note` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.c_singaporeE3.rows[].window` | string | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.c_singaporeE3.rows[].answersSearched` | number | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.c_singaporeE3.rows[].codedNamings` | number | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.c_singaporeE3.rows[].codesByCode.status-correct` | number | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.c_singaporeE3.rows[].codesByCode.status-absent` | number | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.c_singaporeE3.rows[].codesByCode.status-wrong` | number | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item11E1Corrections.c_singaporeE3.outOfScopeContext` | number | E1 corrections: latency as an interval set by the cycle schedule; the two Gemini answers stating the OCTANE result missed by the text route; Singapore E3 namings by window |
| `.item12PageP61.status` | string | post-registration, descriptive and post hoc; an association, not a causal estimate |
| `.item12PageP61.page.id` | string | page P61: opaque id, neutral type, page date to the month, provenance of the typing, first citation in the series; its domain, URL and company are never written |
| `.item12PageP61.page.type` | string | page P61: opaque id, neutral type, page date to the month, provenance of the typing, first citation in the series; its domain, URL and company are never written |
| `.item12PageP61.page.pageDate` | string | page P61: opaque id, neutral type, page date to the month, provenance of the typing, first citation in the series; its domain, URL and company are never written |
| `.item12PageP61.page.provenance` | string | page P61: opaque id, neutral type, page date to the month, provenance of the typing, first citation in the series; its domain, URL and company are never written |
| `.item12PageP61.page.firstCitedInSeries.date` | string | page P61: opaque id, neutral type, page date to the month, provenance of the typing, first citation in the series; its domain, URL and company are never written |
| `.item12PageP61.page.firstCitedInSeries.window` | string | page P61: opaque id, neutral type, page date to the month, provenance of the typing, first citation in the series; its domain, URL and company are never written |
| `.item12PageP61.page.firstCitedInSeries.engine` | string | page P61: opaque id, neutral type, page date to the month, provenance of the typing, first citation in the series; its domain, URL and company are never written |
| `.item12PageP61.page.never` | string | page P61: opaque id, neutral type, page date to the month, provenance of the typing, first citation in the series; its domain, URL and company are never written |
| `.item12PageP61.a_citationsByEngine[].series` | string | citations of P61 by engine and window |
| `.item12PageP61.a_citationsByEngine[].engine` | string | citations of P61 by engine and window |
| `.item12PageP61.a_citationsByEngine[].window` | string | citations of P61 by engine and window |
| `.item12PageP61.a_citationsByEngine[].answers` | number | citations of P61 by engine and window |
| `.item12PageP61.a_citationsByEngine[].citing` | number | citations of P61 by engine and window |
| `.item12PageP61.a_citationsByEngine[].namedAmongCiting` | number | citations of P61 by engine and window |
| `.item12PageP61.b_contentAtCitationByEcho.rule` | string | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.sentences[].id` | string | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.sentences[].description` | string | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.sentences[].normalisedLength` | number | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.claudeByCycle[].slug` | string | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.claudeByCycle[].date` | string | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.claudeByCycle[].window` | string | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.claudeByCycle[].answers` | number | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.claudeByCycle[].citingP61` | number | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.claudeByCycle[].S1.citing` | number | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.claudeByCycle[].S1.notCiting` | number | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.claudeByCycle[].S2.citing` | number | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.claudeByCycle[].S2.notCiting` | number | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.claudeByCycle[].either.citing` | number | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.claudeByCycle[].either.notCiting` | number | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.claudeByCycle[].eitherStrong.citing` | number | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.claudeByCycle[].eitherStrong.notCiting` | number | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.claudeByCycle[].eitherNearFull.citing` | number | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.claudeByCycle[].eitherNearFull.notCiting` | number | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.claudeByCycle[].longestNotCiting` | number | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.preCongressEchoes[].date` | string | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.preCongressEchoes[].window` | string | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.preCongressEchoes[].engine` | string | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.preCongressEchoes[].sentence` | string | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.preCongressEchoes[].matchLength` | number | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.preCongressEchoes[].sentenceLength` | number | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.preCongressEchoes[].sentenceOffset` | number | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.preCongressEchoes[].citesP61` | boolean | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.byEngineAndWindow[].engine` | string | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.byEngineAndWindow[].window` | string | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.byEngineAndWindow[].answers` | number | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.byEngineAndWindow[].echo40` | number | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.byEngineAndWindow[].echo80` | number | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.b_contentAtCitationByEcho.byEngineAndWindow[].echo40CitingP61` | number | content at citation by echo: the two sentences named by neutral labels (S1, S2) with their normalised lengths; per cycle and per engine and window, answers sharing ≥ 40 (and ≥ 80) normalised characters with a sentence, citing and not citing the page; pre-congress echoes |
| `.item12PageP61.c_whereClaudesRiseSits.note` | string | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byWindow[].window` | string | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byWindow[].citingP61.k` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byWindow[].citingP61.n` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byWindow[].citingP61.rate` | null or number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byWindow[].notCiting.k` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byWindow[].notCiting.n` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byWindow[].notCiting.rate` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byCycle[].slug` | string | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byCycle[].date` | string | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byCycle[].window` | string | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byCycle[].citingP61.k` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byCycle[].citingP61.n` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byCycle[].citingP61.rate` | null or number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byCycle[].notCiting.k` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byCycle[].notCiting.n` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byCycle[].notCiting.rate` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byQuestion[].question` | string | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byQuestion[].before.k` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byQuestion[].before.n` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byQuestion[].before.rate` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byQuestion[].{window}.k` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byQuestion[].{window}.n` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byQuestion[].{window}.rate` | number or null | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byQuestion[].afterCitingP61.k` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byQuestion[].afterCitingP61.n` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byQuestion[].afterCitingP61.rate` | number or null | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byQuestion[].afterNotCiting.k` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byQuestion[].afterNotCiting.n` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byQuestion[].afterNotCiting.rate` | number or null | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.byQuestion[].drewP61InAfter` | boolean | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.drewP61.questions` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.drewP61.before.k` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.drewP61.before.n` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.drewP61.before.rate` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.drewP61.{window}.k` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.drewP61.{window}.n` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.drewP61.{window}.rate` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.drewP61.afterCitingP61.k` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.drewP61.afterCitingP61.n` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.drewP61.afterCitingP61.rate` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.drewP61.afterNotCiting.k` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.drewP61.afterNotCiting.n` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.drewP61.afterNotCiting.rate` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.drewP61.questionsWithBeforeRateAtLeastHalf` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.didNotDrawP61.questions` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.didNotDrawP61.before.k` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.didNotDrawP61.before.n` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.didNotDrawP61.before.rate` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.didNotDrawP61.{window}.k` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.didNotDrawP61.{window}.n` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.didNotDrawP61.{window}.rate` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.didNotDrawP61.afterCitingP61.k` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.didNotDrawP61.afterCitingP61.n` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.didNotDrawP61.afterCitingP61.rate` | null | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.didNotDrawP61.afterNotCiting.k` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.didNotDrawP61.afterNotCiting.n` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.didNotDrawP61.afterNotCiting.rate` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.didNotDrawP61.questionsWithBeforeRateAtLeastHalf` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.questionsNotAskedInAfterWindow` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.groups.questionsOnlyInAfterWindow` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.decomposition.rule` | string | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.decomposition.afterAnswersWithBase` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.decomposition.afterAnswersWithoutBase` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.decomposition.excessNamingsCitingP61` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.decomposition.excessNamingsNotCiting` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.decomposition.excessNamingsNotCitingInQuestionsThatDrewP61` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.decomposition.excessNamingsNotCitingInOtherQuestions` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.decomposition.totalExcessNamings` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.decomposition.totalExcessPp` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.c_whereClaudesRiseSits.decomposition.shareOfExcessFromCitingAnswers` | number | where Claude's UAE rise sits: governed naming by window and cycle for answers citing P61 and not; per question (opaque Q-numbers keyed by question text, not the package's prompt identifiers); groups; and the excess-naming decomposition |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.citationsOfP61ByEngine[]` | (empty) | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.citesP61[]` | (empty) | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.naming.2026-10-06.n` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.naming.2026-10-06.registered` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.naming.2026-10-06.brandOnly` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.naming.2026-10-06.moleculeOnly` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.naming.2026-10-07.n` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.naming.2026-10-07.registered` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.naming.2026-10-07.brandOnly` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.naming.2026-10-07.moleculeOnly` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.naming.preStep4And6Oct.n` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.naming.preStep4And6Oct.registered` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.naming.preStep4And6Oct.brandOnly` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.naming.preStep4And6Oct.moleculeOnly` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.naming.postStep7To11Oct.n` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.naming.postStep7To11Oct.registered` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.naming.postStep7To11Oct.brandOnly` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.naming.postStep7To11Oct.moleculeOnly` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.step6To7Oct.registeredPp` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.step6To7Oct.brandOnlyPp` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.step6To7Oct.moleculeOnlyPp` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.stepPreToPost.registeredPp` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.stepPreToPost.brandOnlyPp` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.stepPreToPost.moleculeOnlyPp` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.pairedByQuestion6To7Oct.pairs` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.pairedByQuestion6To7Oct.flipsUp` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.pairedByQuestion6To7Oct.flipsUpBrandNamed` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.pairedByQuestion6To7Oct.flipsUpMoleculeOnly` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.pairedByQuestion6To7Oct.flipsDown` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.groupsCitedOn7Oct[].group` | string | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.groupsCitedOn7Oct[].flipsUpCiting` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.groupsCitedOn7Oct[].flipsUp` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.groupsCitedOn7Oct[].shareOfTurnedCiting` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.groupsCitedOn7Oct[].shareOfAll7Citing` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.groupsCitedOn7Oct[].answers7Citing` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.groupsCitedOn7Oct[].named7AmongCiting` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.groupsCitedOn7Oct[].answers7NotCiting` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.groupsCitedOn7Oct[].named7AmongNotCiting` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.topNewPagesAmongFlips[].page` | string | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.topNewPagesAmongFlips[].domain` | string | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.topNewPagesAmongFlips[].group` | string | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.topNewPagesAmongFlips[].flipsUpCiting` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.topNewPagesAmongFlips[].answers7Citing` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.marketResearchAssociation.preStep.citingMarketResearch.k` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.marketResearchAssociation.preStep.citingMarketResearch.n` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.marketResearchAssociation.preStep.citingMarketResearch.rate` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.marketResearchAssociation.preStep.notCiting.k` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.marketResearchAssociation.preStep.notCiting.n` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.marketResearchAssociation.preStep.notCiting.rate` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.marketResearchAssociation.postStep.citingMarketResearch.k` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.marketResearchAssociation.postStep.citingMarketResearch.n` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.marketResearchAssociation.postStep.citingMarketResearch.rate` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.marketResearchAssociation.postStep.notCiting.k` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.marketResearchAssociation.postStep.notCiting.n` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.d_otherEnginesAndPerplexityStep.{engine}.marketResearchAssociation.postStep.notCiting.rate` | number | other engines' citations of P61 and Perplexity's 7 Oct step, with the question-paired flips from 6 to 7 Oct |
| `.item12PageP61.e_e4Link.answersEchoingS2` | number | the E4 link: the efficacy figure in answers echoing S2 |
| `.item12PageP61.e_e4Link.byEngine.{engine}` | number | the E4 link: the efficacy figure in answers echoing S2 |
| `.item12PageP61.e_e4Link.echoingS2CitingP61` | number | the E4 link: the efficacy figure in answers echoing S2 |
| `.item12PageP61.e_e4Link.withTheJointClassFigure` | number | the E4 link: the efficacy figure in answers echoing S2 |
| `.item12PageP61.e_e4Link.scriptedCategory.label` | number | the E4 link: the efficacy figure in answers echoing S2 |
| `.item12PageP61.e_e4Link.handCheckVerdicts.excluded` | number | the E4 link: the efficacy figure in answers echoing S2 |
| `.item12PageP61.e_e4Link.handCheckReasons.P3` | number | the E4 link: the efficacy figure in answers echoing S2 |
| `.item12PageP61.e_e4Link.allHandCheckedExclusionsOfThisFigureP3` | number | the E4 link: the efficacy figure in answers echoing S2 |
| `.item12PageP61.e_e4Link.ofWhichInAnswersEchoingS2` | number | the E4 link: the efficacy figure in answers echoing S2 |
| `.item12PageP61.e_e4Link.reasonP3` | string | the E4 link: the efficacy figure in answers echoing S2 |

## `robustness/easd-2026-robustness.md`

- **Role:** robustness-md. The D24 tables as printed by the same run.
- **Endpoint:** POST-REGISTRATION ROBUSTNESS (D24)
- **Source:** `docs/evidence/easd-2026-robustness.md` (frozen); transformed (see REDACTION-LOG.md)
- **Structure:** POST-REGISTRATION ROBUSTNESS (D24): one section per item (1 registered §5 analyses reported late, 2 band readings, 3 cycle-level change, 4 per-cycle view, 5 corrected self-check, 6 brand-only naming, 7 search rate, 8 attribution sensitivity, 9 Saudi use context, 10 per-answer table, 11 E1 corrections, 12 page P61) and the script's own disclosure scan.

## `robustness/easd-2026-saudi-e3i-use-context.json`

- **Role:** saudi-e3i-use-context. Model-coded type 1 use context (glycaemic, weight-management, unspecified) for every coded type 1 use naming; same model family, not validation.
- **Endpoint:** POST-REGISTRATION ROBUSTNESS (D24) item 9: Saudi E3-i use-context recode (model-coded sensitivity)
- **Source:** `docs/evidence/easd-2026-saudi-e3i-use-context.json` (frozen); transformed (see REDACTION-LOG.md)

| Field | Type | Meaning, units and allowed values |
|---|---|---|
| `.label` | string | POST-REGISTRATION ROBUSTNESS (D24) label |
| `.deviation` | string | D24 item 9 |
| `.status` | string | model-coded sensitivity; same model family; not human coding and not validation |
| `.codedAt` | string | date coded |
| `.population` | string | which namings are coded (every sampled E3-i type 1 use naming) |
| `.procedure` | string | how each naming was read |
| `.rules[]` | string | the context rules R1 glycaemic, R2 weight-management, R3 unspecified (verbatim) |
| `.codes.{sha256}[].agent` | string | molecule (semaglutide \| tirzepatide \| sotagliflozin) |
| `.codes.{sha256}[].window` | string | window: before-v4 \| before-v5 \| during \| after (protocol §3) |
| `.codes.{sha256}[].registeredCode` | string | the registered E3-i code of the naming (indication-stated \| indication-absent \| indication-wrong) |
| `.codes.{sha256}[].context` | string | glycaemic \| weight-management \| unspecified |
| `.codes.{sha256}[].weightAlsoStated` | boolean | a weight purpose is also stated (true \| false) |
| `.codes.{sha256}[].populationOverweightOrObese` | boolean | the use is framed for people with type 1 diabetes and overweight or obesity (true \| false) |
| `.codes.{sha256}[].note` | string | coder's note |
| `.codes.{sha256}[].span` | string | the coded answer span (≤ about 240 characters); a marker for Gemini answers |
| `.codes.{sha256}[].spanSha256` | string | sha256 of the withheld Gemini answer span (present only where the span is withheld, R1m; the Gemini reporting policy) |

## `saudi/easd-2026-t1d-agent-set.md`

- **Role:** t1d-agent-set. The type 1 diabetes agent set, its programme sources and membership rule, committed before coding.
- **Endpoint:** Saudi E3 and E3-i membership (D12)
- **Source:** `docs/evidence/easd-2026-t1d-agent-set.md` (frozen); byte-identical to the source
- **Structure:** membership rule, one entry per agent with its programme source and regulator records, E3-i molecules, and non-programme agents counted for context.

## `second-pass/easd-2026-second-pass-builders.py`

- **Role:** second-pass-builders. The two worksheet builders of the blind pass, as run (sanitised).
- **Endpoint:** Second-pass sampling and blinding
- **Source:** `docs/evidence/easd-2026-second-pass-builders.py` (frozen); transformed (see REDACTION-LOG.md)
- **Structure:** source code; its header comment states what it reads, writes and never does, and its usage.

## `second-pass/easd-2026-second-pass.json`

- **Role:** second-pass. The blind second coding pass by a separate instance of the same model family, with adjudications.
- **Endpoint:** E3, E5, E3-i reliability (robustness, D16)
- **Source:** `docs/evidence/easd-2026-second-pass.json` (frozen); transformed (see REDACTION-LOG.md)

| Field | Type | Meaning, units and allowed values |
|---|---|---|
| `.what` | string | description of the pass |
| `.seed` | number | random seed (20261011) |
| `.stats.{key}.items` | number | items per endpoint (E3 \| E5 \| E3-i) |
| `.stats.{key}.agreement` | number | raw agreement, proportion |
| `.stats.{key}.cohensKappa` | number | Cohen's kappa |
| `.stats.{key}.disagreements` | number | items where the codes differ |
| `.adoptedFromSecondPass` | number | adjudications that adopted the second code |
| `.keptFirstCode` | number | adjudications that kept the first code |
| `.items[].id` | number | blind item number |
| `.items[].endpoint` | string | E3 \| E5 \| E3-i |
| `.items[].sha256` | string | sha256 of the stored answer capture; links rows across files and to the held pull |
| `.items[].agent` | null or string | agent or molecule; null for E5 |
| `.items[].firstCoder` | string | first code |
| `.items[].secondPass` | string | blind second code |
| `.items[].secondPassNote` | string | second coder's note |
| `.items[].adjudicated` | string | final code |
| `.items[].adjudication` | null or string | adjudication note; null when the codes agreed |
| `.items[].batch` | string | empty for batch 1; top-up 2026-10-11 |

## `validation/human-validation-protocol.md`

- **Role:** human-validation-protocol. The protocol of the optional blinded human validation: sample, blinding, analysis and reporting.
- **Endpoint:** Optional blinded human check of naming and interpretive codes (validation policy)
- **Source:** `docs/research/easd-2026/human-validation-protocol.md` (frozen); transformed (see REDACTION-LOG.md)
- **Structure:** §1 what the check is and is not; §2 inputs; §3 the sample (Part A naming, 100; Part B interpretive judgements, 50; seed 20261012); §4 blinding; §5 analysis and reporting; §6 limitations; §7 files and reproduction.

## `validation/human-validation-results.md`

- **Role:** human-validation-results. The results of the blinded human check: Part B agreement by endpoint and sample with Wilson intervals and kappa, the cross-tabulations, every disagreement and the reader's notes. Item IDs, codes, engines, windows and answer sha prefixes only; no question or answer text.
- **Endpoint:** Blinded human validation of the interpretive coding (Part B, 50 items); a single reader, the author; not independent validation
- **Source:** `docs/research/easd-2026/human-validation-results.md` (frozen); transformed (see REDACTION-LOG.md)
- **Structure:** provenance (input hashes, scope); Part A (not coded, by the author's decision); Part B agreement by endpoint and sample, with Wilson 95% intervals and Cohen's kappa; cross-tabulations; every disagreement (item, sample, stratum, market, engine, window, both codes, answer sha256 prefix); the reader's notes on agreeing items; how to read the figures.

## `verification/easd-2026-verify.txt`

- **Role:** verify. The verifier's frozen output (D21 item 7).
- **Endpoint:** Coding consistency (all coded endpoints)
- **Source:** `docs/evidence/computed/easd-2026-verify.txt` (frozen); transformed (see REDACTION-LOG.md)
- **Structure:** one line per check: PASS, FAIL or INFO and the check; the last line states the result.