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

Safety Evaluation on HarmBench Std

80.8Refusal Rate

ALIGNBEAM

62.80867.47972.1576.821Jun 10, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
80.889.8
2026.06
79.691
2026.06
79.691.6
2026.06
78.488.6
2026.06
76.489.2
2026.06
63.585