Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Safety Evaluation on HarmBench Ctx

80.6Refusal Rate

ALIGNBEAM

60.52865.73970.9576.161Jun 10, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
80.690.3
2026.06
7980.7
2026.06
72.385.5
2026.06
7190.3
2026.06
69.487.1
2026.06
61.379