AI/ML●●Solid
DocForge – Multi-Agent RAG That Fact-Checks Its Own Answers
Multi-agent fact-checking loop, but RAG hallucination fixes are table stakes now.
Big BrainShip It
toheed11
114mo ago
Back to browse
GitHub Repository
A memory layer that tracks evidence, claims, and decisions to make multi-turn LLM judges and reviewer agents more inspectable and stable.
2 starsPython
I made a small helper for checking model-graded answers
by ML0037·Jun 14, 2026·2 points·0 comments
AI Analysis
●●●BangerBig BrainSolve My ProblemDark Horse
Structurally verifies LLM judge reasoning instead of paying for a second model check.
Strengths
- •Structural verification flags ignored references without needing a second LLM judge call.
- •Cites specific research papers on judge bias to justify the design decisions clearly.
- •Local CLI viewer allows inspecting flagged runs immediately without cloud dashboard setup.
Weaknesses
- •Web dashboard mentioned in README appears unfinished compared to the local CLI viewer.
- •Requires modifying eval prompts to fit the claim-evidence structure, adding integration friction.
Category
Target Audience
AI researchers and ML engineers running LLM evaluations
Similar To
LangSmith · Arize Phoenix · DeepEval
Post Description
I made this while checking model graded answer and helped me to check the odd cases by hand. Not sure if it’s useful to anyone else.
TL;DR: it breaks an LLM judge run into claims->evidence->verdicts and flags when a verdict is not supported by the evidence, so i can check it manually.
Similar Projects
AI/ML●●Solid
LLM Debate Benchmark
Side-swapped debate matchups expose model weaknesses standard benchmarks miss.
Big BrainDark Horse
zone411
932mo ago
Productivity●Mid
Climbing Grade Converter
Yet another grade converter when several already exist.
Cozy
citrus1330
3013d ago
AI/ML●●●Banger
Auto LLM Ranker – Describe a task in English and get ranked models
Task-specific LLM benchmarking beats generic leaderboards that ignore your actual workload.
Big BrainDark HorseZero to One
gauravvij137
303mo ago
AI/ML●●Solid
Signals – finding the most informative agent traces without LLM judges
Heuristic signals triage agent traces 1.52x more efficiently than random sampling.
Big BrainNiche Gem
sparacha
302mo ago
Developer Tools●●Solid
Yardstiq – Compare LLM outputs side-by-side in your terminal
One-command model comparison with real-time streaming and performance metrics beats tab-switching.
Solve My ProblemSlickCrowd Pleaser
stanleycyang
203mo ago