Mysticbirdie/hallucination-elimination-benchmark

View on GitHub ↗ ★ 7 Python

Multi-tier benchmark: Cultural grounding + Triad Engine eliminates LLM hallucination across Claude 4.6, GPT-5.2, Mistral 7B, Gemini 2.5 Pro. Raw 15-58% → 95-100% accuracy on 222 adversarial QA pairs (Ancient Rome 110 CE). Novel topological paradox detection (F1=0.939, zero-shot). Model-agnostic, in production.

3 mentions on Hacker News by 1 person: 2 stories, 1 comment · first 2026-02-25 · latest 2026-02-26

Mentions per year 3 mentions · busiest year 2026 (3)

20082026

All Stories Comments Newest Top Oldest

▲ 2
Claude Code Mexico breach: training safety failed ground truth layer
story by MysticBirdie · 2026-02-26 · 1 comment
▲ 2
Show HN: 100% LLM accuracy–no fine-tuning, JSON only
story by MysticBirdie · 2026-02-25 · 2 comments