Technology
BenchMIRT
StableFollow
01
Recent changes
- BenchMIRT: A New Method for Auditing LLM BenchmarksBenchMIRT is a new method for auditing large language model (LLM) benchmarks at the level of individual prompts, revealing insights into existing benchmarks and suggesting more efficient evaluation methods.StableTechnology1 sources · 1 primary · 2 evidence2026-09-02 05:39