failure-transparent-agents
Benchmarking whether LLM agents admit tool failure instead of hallucinating success — 100 tasks, 1,800 responses, deterministic simulator, reproducible harness.
Back to catalogSigned-out saves stay in this browser. Sign in to use account saves.
Loading catalog…카탈로그를 불러오는 중…
RECENT CHANGE
What changed
Loading recent changes…
Per-day rates divide a measured star count by the days it covers: the interval between two snapshots for Rising fastest, the age of the repository for New and fast. They are not fixed 24-hour changes, and stars show interest, not quality.