safeagents judge bench v2 launches leaderboard on Hugging Face
safeagents judge bench v2 is live on Hugging Face, featuring 661 real agent tool calls labeled allow/block. Cloudflare's clef-flash tops the leaderboard with 0.893 balanced accuracy, showing small models can serve as agent safety judges.
2026-11-10 ~ 2026-11-10 · 3 related posts
- safeagents judge bench lands on HF: official leaderboard for decision-model safety judges — yb2698 · 2026-11-10
- safeagents judge bench v2: 661 labeled agent tool calls to score safety judge models — yb2698 · 2026-11-10
- safeagents judge bench v2 leaderboard: Cloudflare clef-flash tops at 0.893 — yb2698 · 2026-11-10