docs(taxonomy): #498 AIF generative layer pilot — GATED proposal (18-fallacy, gpt-5.5, anti-fab verified) - #505
Merged
Conversation
Adds a GATED proposal (no production CSV change) for the AIF generative layer: exception-structure annotation for Fallacies nodes that have no AIF analogue. Gap & reformulation (per ai-01): only 42/1408 Fallacies carry a direct AIF_skosDirectRef. AIF has ~20 native schemes, so the value is NOT mapping to its limited corpus but the GENERATIVE exception structure — each fallacy = a defeasible exception to a Walton scheme that evades a critical question (CQ). A latin name is 0-shot consumable; the « exception to scheme S / violated CQ Q » mechanism is not. Contents: - 18-fallacy pilot (depth-3, NO prior AIF, ~2-3 per family across all 7): each annotated with canonical Walton scheme + violated CQ (FR) + exception mechanism + justification. Several textbook-classic mappings (Consequences, Analogy, Verbal Classification, Bias, Cause to Effect). - Method: gpt-5.5 via /v1/responses + reasoning.effort:low (cluster's proven reliable-content method, ~68 reasoning tokens/call), grounded in a 24-scheme canonical Walton catalog. - Anti-fabrication (3 layers): scheme ∈ catalog, PK ∈ real 1408-row CSV, all 18 confirmed no-AIF in source. 0 violations. 12 distinct schemes across 18 fallacies (healthy distribution). GATED: ai-01 reviews structure, jsboige validates content. No Fallacies CSV edit, no OWL/consumer change until approval. Pilot = calibration exemplar for the ~1366 no-AIF scale-up; feeds EPITA Restitution #1134. Refs #498 Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Collaborator
|
[NanoClaw] LGTM — GATED proposal, 2 files (+109/-0) Smart reformulation of the AIF gap: instead of chasing impossible corpus-node coverage (~20 native AIF schemes for 1408 fallacies), documents the defeasible exception structure (Walton scheme + violated CQ). 18-fallacy pilot with 12 distinct schemes — healthy distribution, no over-concentration. Anti-fab solid: 3-layer verify (scheme catalog, real PKs, no-prior-AIF check), 0 violations. Gate boundaries clear (no production CSV, no OWL, no consumer changes). No concerns. |
jsboige
marked this pull request as ready for review
June 17, 2026 00:02
42 tasks
jsboige
added a commit
that referenced
this pull request
Jun 17, 2026
Encodes the manual anti-fabrication passes done per taxonomy scale-up phase (pilots #503/#505/#508, phase-1 #509/#510) into a stdlib-only Python tool so each new proposition CSV can be validated automatically before its PR. Dispatched by ai-01 (msg-20260617T051231); built by worker po-2024. Why a script, not a C# test: the docs/taxonomy/ propositions have HETEROGENEOUS schemas (virtues / AIF-triple / cross-links) that are not C# entities, and the generation+anti-fab tooling for this lane is Python. A stdlib-only script adds no runtime the repo does not already imply, and accepts arbitrary CSV paths so a worker validates a phase file on its own branch before opening the PR. Checks (per detected schema kind): - (a) HARD PK membership — every *_pk / opposed PK / family PK / source+target PK is a real PK in the corresponding corpus CSV; family PKs are depth 1. - (b) HARD link_type in the 8 crossLink_* types. - (c) WARN Walton scheme in the union used by validated pilots (drift detection). - (d) HARD attack_type in {undermine, undercut, rebut}. - (e) WARN attacked_component coherent with attack_type (data-grounded map). - (f) HARD symmetric links: the reverse edge is encoded (symmetric flag = source of truth, not a type-set heuristic). Anti-fab design: no machine-readable canonical "Walton 24" exists in the repo (corpus AIF_skosDirectRef holds AIF node IDs, not scheme names; generation was ad-hoc). Hardcoding Walton from memory risks false fail/pass. So check (c) is a WARNING built from the union of scheme names used by merged validated pilots — catches drift honestly without fabricating a canonical list. DoD: 0 HARD violations across all 5 existing CSVs (497-pilot-crosslinks, 498-pilot-annotations, 499-pilot-annotations on master + 498-scaleup-phase1 #509, 499-scaleup-phase1 #510 from their branches). Negative test: deliberately corrupted CSVs trigger every HARD check (bad PK, bad link type, wrong family depth, bad attack type, missing reciprocal edge) — proven, not assumed. Gate-safe: pure tooling, 0 prod change (no CSV, no config, no workflow). Contributes to the taxonomy scale-up lane (#497/#498/#499). Refs #509/#510. Co-authored-by: Your <your.email@example.com> Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
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
2 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.
What
GATED proposal for #498 — AIF generative layer. Per ai-01's reformulation, the value is NOT mapping Fallacies to AIF's ~20-node corpus (only 42/1408 carry a direct
AIF_skosDirectRef), but the generative exception structure: each fallacy = a defeasible exception to a Walton scheme that evades one of its critical questions (CQ). This PR proposes the framing + demonstrates it on an 18-fallacy pilot, with a 3-layer anti-fabrication guarantee. No production Fallacies CSV is changed — proposal for jsboige validation before scale-up.Why
A latin name (« argumentum ad consequentiam ») is consumable by any LLM 0-shot; the « exception to scheme S / violated CQ Q » mechanism is not — and it's the layer EPITA's formal verification and the Restitution Epic #1134 actually need to narrate why a fallacy fails. ai-01's cross-workspace synchro (EPITA × Argumentum × CoursIA) frames this as the readable debouché of the formal layer.
The pilot (18 fallacies — all 7 families, all currently NO-AIF)
Each fallacy → canonical Walton scheme it's an exception to + the violated critical question (FR) + exception mechanism + justification. Example rows:
Several are textbook-classic mappings. Full 18 rows + mechanisms in
docs/taxonomy/498-aif-generative-pilot.md+498-pilot-annotations.csv.Anti-fabrication guarantee (3 layers)
walton_scheme∈ catalog, (2)fallacy_pk∈ real 1408-row CSV, (3) all 18 confirmed no prior AIF in source CSV (generative-layer claim holds).Method
gpt-5.5 via
/v1/responses+reasoning:{effort:"low"}— the cluster's proven method for reliable content (chat/completions burns the budget in reasoning_tokens and can return empty; effort:low yielded clean JSON at ~68 reasoning tokens/call, validated this tick with a one-fallacy smoke before the batch). One call per fallacy.Gate boundaries (what this does NOT do)
Argumentum Fallacies - Taxonomy.csv.For reviewers
Marked draft until jsboige content review. Companion to #503 (#499 Virtues parity) — both are the taxonomy-generative lane.
Refs #498
🤖 Generated with Claude Code