/TruthEval

A curated collection of challenging statements on sensitive topics for LLM benchmarking. Designed to distinguish LLMs' abilities from their stochastic nature.

No issues in this repository yet.