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
20082026
-
▲ 2
Claude Code Mexico breach: training safety failed ground truth layerstory by MysticBirdie · 2026-02-26 · 1 comment
-
Follow-up: we ran adversarial chaining tests after a few questions about multi-turn behavior. Two chain types (Gemini 2.0 Flash, same model for both): Depth scaling (chains of depth 3, 5, 10, 20, 50 — 91 total steps): no drift in either…
comment by MysticBirdie · 2026-02-25 · in Show HN: 100% LLM accuracy–no fine-tuning, JSON only -
▲ 2
Show HN: 100% LLM accuracy–no fine-tuning, JSON onlystory by MysticBirdie · 2026-02-25 · 2 comments