data(virtues): v2 multilang cascade FP gpt-5.5 (228 cells / 211 records, 0 leak) - #385
Closed
jsboige wants to merge 4 commits into
Closed
data(virtues): v2 multilang cascade FP gpt-5.5 (228 cells / 211 records, 0 leak)#385jsboige wants to merge 4 commits into
jsboige wants to merge 4 commits into
Conversation
…cies prompts Phase 4 quality remediation, driven by sample-based evaluation (seeds 42/123/7): - Scenarii PK 4.2.8 PT: bem -agua + EN mistranslation (bitter evening) - Virtues pk 68: EN/RU/PT all re-conceptualised title (Assessment of independence vs FR Independance des evenements) - Fallacies PK 51: EN/PT semantic drift from desc_fr (closer to James than to Anthropomorphisme) Changes: 1. New Scenarii PT refine task (gpt-5.5, full pass) - FieldsToInclude: FR (locked) + EN (mirror) + PT (editable) - FieldsToUpdate: title_pt, context_pt, issue_pt, suggestion_pt - Targets residual non-mechanical PT artefacts after deterministic cleanup (#383) - TakeChunkNb=-1, MaxDegreeOfParallelismWebService=4 2. Enhanced PromptVirtuesCascadeDriftUser - New criterion: title re-conceptualisation (align to title_fr concept) 3. Enhanced PromptFallaciesCascadeDriftUser - New criterion: structural semantic drift (align to desc_fr taxonomy) Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
…rds) Full-pass PT remediation following deterministic enclitic cleanup (#383). Targets residual non-mechanical artefacts via dedicated PT refine prompt. Stats: - title_pt: 56 cells - context_pt: 94 cells - issue_pt: 94 cells - suggestion_pt: 100 cells - Total: 344 cells across 129/167 records (77%) - Scope leak: 0 (no non-PT field touched) Sample improvements: - PK 1.1.1 "ditador da vida" -> "ditador vitalico" (calque fix) - PK 1.1.1 "dissuadi -lo" -> "dissuadi-lo" (residual enclitic + period) - PK 1.1.2 "fazer um aliado" -> "tornar-se uma aliada" (idiomatic) - PK 1.1.3 "jogos de circo" -> "jogos circenses" (lexical precision) - PK 1.1.3 "Mmm, esta e uma escolha para meus gatinhos" -> "Eis um prato de primeira para os meus leoes" (character voice for Caligula) Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
…/ 211 records) Virtues taxonomy quality remediation — v2 prompt with title re-conceptualisation criterion added. Full pass over 223 records, ChunkSize=4, parallelism=4 → 1:53 wallclock. Distribution (8 langs × 3 fields, 0 FR leak, 0 non-target leak): - remark_en: 157 cells (69%) — most drift-prone field, concretised with examples - remark_ru: 20 cells — concision + formal Вы alignment - description_en: 10 cells - remark_ar: 9 cells - remark_zh: 5 cells - remark_fa: 5 cells - title_en: 4 cells (incl PK 68 "Assessment of independence" → "Independence of events" — exact prompt target hit) - remark_es: 3 cells - remark_pt: 2 cells - title/desc per lang: 1-2 cells each across ar/zh/fa Sample quality leap (PK 38 "Vocabulaire approprié"): OLD: "Care in vocabulary is crucial for effective and honest communication. Words chosen should be..." NEW: "Prefer 'peer-reviewed publication' to 'article' when referring to scientifically validated writings." Sample concision (PK 13 "Sources crédibles" RU): OLD: 3-sentence essay NEW: "Аргумент становится сильнее, когда Вы опираете его на источники, признанные за их точность и объективность." Config: Scenarii PT refine task disabled (FP completed via PR #384); Virtues cascade EN-only renamed/extended to "v2 multilang" (7 langs × 3 fields), TakeChunkNb=-1, parallelism=4, ChunkSize=4. Both tasks remain Enabled=false in committed state. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
…fault) Mirrors the pattern of all existing cascade tasks (Rules/Scenarii/Virtues/Fallacies cascade EN-only) — kept in config for replication but Enabled=false to prevent accidental re-runs and quota burn. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
clusterManager-Myia
approved these changes
May 29, 2026
clusterManager-Myia
left a comment
clusterManager-Myia
left a comment
Collaborator
There was a problem hiding this comment.
[Hermes] — APPROVED
PR: data(virtues): v2 multilang cascade FP gpt-5.5 (228 cells / 211 records, 0 leak)
Checklist
- Security scan: false positives only (CSV narrative text, not credentials)
- Impact cross-repo: none (data-only)
- Change type: CSV data — 228 cells multilang regenerated via gpt-5.5 cascade
Verification
- +224/-224 in Virtues taxonomy CSV
- desc_fr preserved as source, EN/RU/PT cascaded
- Consistent with #384 (same pipeline, same model)
- Title confirms 0 leak
No concerns.
6 tasks
clusterManager-Myia
left a comment
clusterManager-Myia
left a comment
Collaborator
There was a problem hiding this comment.
[Hermes] — COMMENT_WITH_CONCERNS\n\nRisk: MEDIUM-HIGH (bulk multilang data + config, +476/-344, 7 files)\n\nChecklist:\n1. Security scan: CLEAN\n2. Cross-repo impact: Virtues taxonomy + Scenarii cards\n3. Data changes: 228 cells / 211 records\n\nConcerns:\n- mergeable: CONFLICTING — rebase before merge.\n- 7 files including both CSV datasets AND DatasetUpdaterRootConfig.cs — the config changes overlap with #384. Merge ordering matters; recommend merging #384 first.\n- "0 leak" claim in the title: please provide the validation script output or cell count diff that confirms this.
Contributor
Author
|
jsboige
added a commit
that referenced
this pull request
May 29, 2026
…PRs (#395) Cherry-pick the two v2 prompt-engineering enhancements developed in the now-stale stacked PRs #384/#385/#386, decoupled from their conflicting data payloads: - PromptFallaciesCascadeDriftUser: add 'Drift sémantique structurel' criterion (translated desc_* carries a concept different from desc_fr → realign on FR, which is authoritative). - PromptVirtuesCascadeDriftUser: add 'Re-conceptualisation des titres' criterion (translated title drifted to a different concept than title_fr → realign on the exact FR concept). Both files are already referenced by the cascade task configs merged in #370; tasks remain Enabled=false. Additive, no data touched, no conflict. Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This was referenced May 29, 2026
jsboige
added a commit
that referenced
this pull request
May 30, 2026
…cascade (#396) Part A of two-population strategy from ai-01 dispatch (FleetView 18:55). This commit captures the 0-API-cost salvage of FR-stable cells from the 4 quarantined cascade PRs. The remaining 61 PKs / 61 cells where FR was modified post-cascade (AUDIT_FR scope) will be re-cascaded with gpt-5.5 in Part B (separate PR). === Cell salvage (1,476 cells, 0 API calls) === * Virtues (Argumentum Virtues - Taxonomy.csv): 211 cells from #385 - Source: feat/virtues-cascade-v2-multilang (HEAD dc6c486) - Method: full-file checkout (FR untouched, drift = 0) * Fallacies (Argumentum Fallacies - Taxonomy.csv): 960 cells - 942 from #386 (feat/fallacies-cascade-v2-multilang-results, ec4056f) * Binary-safe row-byte transplant (preserves source quoting + CRLF) * Skipped 61 AUDIT_FR PKs (FR modified post-base in master) → Part B * Invariant proven: 0 FR cells touched in #386 vs base 1811afc - 18 from #355 (data/fallacies-desc-realign-40pk, e628994) * 6 orphan PKs (208/281/286/289/730/1074) × 3 cells (desc_en/ru/pt) * Cell-by-cell mapping by column name (header drift on col 33 'latin' → 'latin1' between #355 base and current master) * Scenarii (Argumentum Scenarii - Cards.csv): 300 PT cells - Source: feat/scenarii-pt-llm-refine (HEAD d3c28ad, #384) - Non-regression vs #383 (PR #383 already merged: deterministic enclitique-hyphen + MT-artefact cleanup) - Rule: apply #384 PT value ONLY where master == base for that cell (i.e. #383 cleanup did not touch it). Skipped 44 cells where #383 cleanup wins. - Per-column: title_pt=55, context_pt=90, issue_pt=56, suggestion_pt=99 === Net-new code from #384 === * DatasetUpdater config block (Enabled=false): - "Scenarii PT refine gpt-5.5" task definition - SequentialChunks, ChunkSize=12, gpt-5.5, MaxChildren=8 - Inserted at end of Configs list, after "Fallacies cascade EN-only" * Prompt files (2 net-new): - PromptScenariiPtRefineUser.txt (+63 lines) - PromptScenariiPtRefineAssistant.txt (+8 lines) * EXCLUDED from #384 (already merged via PR #395): - +1 line in PromptFallaciesCascadeDriftUser.txt ("Drift sémantique structurel") - +1 line in PromptVirtuesCascadeDriftUser.txt ("Re-conceptualisation des titres") === Integrity verification === * FR drift vs master HEAD (5cda433): 0 cells across all 3 CSVs * Row counts preserved: Fallacies 1409, Virtues 224, Scenarii 168 * Build: 0 errors (dotnet build, only pre-existing warnings) * CSV format: CRLF + UTF-8 (no BOM on Fallacies/Virtues, BOM on Scenarii) preserved === Files & deltas === * Cards/Fallacies/Argumentum Fallacies - Taxonomy.csv (1306 lines diff) * Cards/Fallacies/Argumentum Virtues - Taxonomy.csv (422 lines diff) * Cards/Scenarii/Argumentum Scenarii - Cards.csv (258 lines diff) * .../DatasetUpdater/DatasetUpdaterRootConfig.cs (+47 lines) * .../Resources/PromptScenariiPtRefineUser.txt (+63 lines, NEW) * .../Resources/PromptScenariiPtRefineAssistant.txt (+8 lines, NEW) Part B (separate PR) will re-cascade the 61 AUDIT_FR PKs × 7 langues (EN/RU/PT/AR/ES/FA/ZH) using gpt-5.5 cascade-drift task to align non-FR translations with the new FR text. Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Closed — superseded by #396 (merged
|
Contributor
Author
Closed — superseded by #396 (merged
|
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Virtues taxonomy v2 multilang cascade — full pass after FR clarity drift (PR #367) with enhanced prompt criteria.
Key prompt enhancement (already merged via PR #384): "Re-conceptualisation des titres" criterion — flag titres traduits qui ont glissé vers un concept différent du
title_fr. Exact target hit: PK 68title_fr="Indépendance des événements"→title_enwas "Assessment of independence" → corrected to "Independence of events".Run
Cells changed (228 across 211/223 records)
Sample quality leaps
PK 38 "Vocabulaire approprié" (concretisation):
PK 13 "Sources crédibles" RU (concision + Вы):
PK 22 "Faits vérifiés" (abstract → actionable):
Config
Virtues cascade v2 multilang gpt-5.5task added (Enabled=false post-FP, defensive)Scenarii PT refine gpt-5.5task disabled (FP completed via PR feat(dataset-updater): Scenarii PT refine + enhanced cascade prompts (v2 quality remediation) #384)Test plan
🤖 Generated with Claude Code