WFGY/ProblemMap/GlobalFixMap/Retrieval/checklists/traceability_gate.md
2025-09-01 16:36:57 +08:00

5.5 KiB
Raw Blame History

Traceability Gate Checklist

Purpose: enforce cite then explain with a minimal, verifiable payload contract.


Required fields per snippet

  • snippet_id unique within the corpus
  • section_id stable anchor for the document
  • source_url resolvable or a content hash if private
  • offsets byte or char range within the source
  • tokens token range used by the model

Contract details:
Retrieval traceability · Data contracts


Schema example

{
  "snippet_id": "sn_9f3a",
  "section_id": "sec_2.3_hydraulics",
  "source_url": "https://example.org/handbook/2024/chap2#sec-3",
  "offsets": {"start": 18342, "end": 18897},
  "tokens": {"start": 1205, "end": 1248}
}

Gate checks

  • Each answer references at least one snippet with all required fields.
  • Offsets resolve back to exactly the cited text.
  • Section anchors stable across builds. Hash or ID diff is logged.
  • Duplicate or near duplicate snippets collapsed before ranking.
  • Formatter and export steps do not rename required fields.
  • On failure, publish is blocked and the pipeline returns a structured error.

Refs: Duplication and near duplicate collapse


Acceptance targets

  • 100 percent of published answers pass the schema check.
  • For the gold set, human spot check of 20 answers finds zero broken citations.
  • ΔS(question, cited_snippet) ≤ 0.45 on average.

Minimal policy snippet

Policy:
- Answer only after citing snippets with {snippet_id, section_id, source_url, offsets, tokens}.
- If no snippet qualifies, return "insufficient evidence" with the top 3 candidate snippets and exit.

🔗 Quick-Start Downloads (60 sec)

Tool Link 3-Step Setup
WFGY 1.0 PDF Engine Paper 1 Download · 2 Upload to your LLM · 3 Ask “Answer using WFGY + <your question>”
TXT OS (plain-text OS) TXTOS.txt 1 Download · 2 Paste into any LLM chat · 3 Type “hello world” — OS boots instantly

🧭 Explore More

Module Description Link
WFGY Core WFGY 2.0 engine is live: full symbolic reasoning architecture and math stack View →
Problem Map 1.0 Initial 16-mode diagnostic and symbolic fix framework View →
Problem Map 2.0 RAG-focused failure tree, modular fixes, and pipelines View →
Semantic Clinic Index Expanded failure catalog: prompt injection, memory bugs, logic drift View →
Semantic Blueprint Layer-based symbolic reasoning & semantic modulations View →
Benchmark vs GPT-5 Stress test GPT-5 with full WFGY reasoning suite View →
🧙‍♂️ Starter Village 🏡 New here? Lost in symbols? Click here and let the wizard guide you through Start →

👑 Early Stargazers: See the Hall of Fame
Engineers, hackers, and open source builders who supported WFGY from day one.

GitHub stars WFGY Engine 2.0 is already unlocked. Star the repo to help others discover it and unlock more on the Unlock Board.

WFGY Main   TXT OS   Blah   Blot   Bloc   Blur   Blow