WFGY/ProblemMap/GlobalFixMap/Language/multilingual_guide.md

7.4 KiB
Raw Permalink Blame History

Multilingual Guide — Guardrails and Fix Patterns

🧭 Quick Return to Map

You are in a sub-page of Language.
To reorient, go back here:

Think of this page as a desk within a ward.
If you need the full triage and all prescriptions, return to the Emergency Room lobby.

A compact field guide to stabilize multilingual RAG across CJK, RTL, mixed scripts, and locale drift. Use this page to check symptoms, apply structural fixes, and verify with measurable targets.


Open these first


Core acceptance

  • ΔS(question, retrieved) ≤ 0.45 across at least 2 languages
  • Coverage ≥ 0.70 for the target section in each language
  • λ remains convergent across three paraphrases in mixed scripts
  • E_resonance stays flat for long bilingual/RTL runs

Common multilingual failure modes

Symptom Likely cause Open this
Retrieval drops snippets when query is in Chinese or Japanese Tokenizer mismatch (no whitespace segmentation) tokenizer_mismatch.md
Citations collapse when Arabic/Hebrew text mixes with English Script directionality conflict script_mixing.md
High similarity but meaning flips across locale Locale analyzer mismatch (stemming / stopwords) locale_drift.md
HyDE/BM25 retrieval different per language Query expansion language bias hyde_multilingual.md

Fix in 60 seconds

  1. Probe with ΔS Run the same question in English and one target language. If ΔS differs by >0.15, suspect tokenization or analyzer mismatch.

  2. Apply λ_observe Paraphrase the query three ways in the non-English language. If λ diverges, enforce schema lock and re-index with language-specific analyzers.

  3. Structural repair


Diagnostic checklist

  • Tokenizer: verify segmentation strategy (whitespace vs character-level)
  • Analyzer: confirm stemming and stopword lists match query language
  • Scripts: normalize Unicode, check RTL/LTR flags
  • Locale drift: run same snippet under two locales, compare ΔS
  • Hybrid retriever: ensure rerankers operate on normalized embeddings

🔗 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

Layer Page What its for
Proof WFGY Recognition Map External citations, integrations, and ecosystem proof
⚙️ Engine WFGY 1.0 Original PDF tension engine and early logic sketch (legacy reference)
⚙️ Engine WFGY 2.0 Production tension kernel for RAG and agent systems
⚙️ Engine WFGY 3.0 TXT based Singularity tension engine (131 S class set)
🗺️ Map Problem Map 1.0 Flagship 16 problem RAG failure taxonomy and fix map
🗺️ Map Problem Map 2.0 Global Debug Card for RAG and agent pipeline diagnosis
🗺️ Map Problem Map 3.0 Global AI troubleshooting atlas and failure pattern map
🧰 App TXT OS .txt semantic OS with fast bootstrap
🧰 App Blah Blah Blah Abstract and paradox Q&A built on TXT OS
🧰 App Blur Blur Blur Text to image generation with semantic control
🏡 Onboarding Starter Village Guided entry point for new users

If this repository helped, starring it improves discovery so more builders can find the docs and tools.
GitHub Repo stars