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

LLM Safety and Informativeness Evaluation on HarmBench

82.5Safety Rate

SafeMoE-XL

11.57229.98648.466.814May 30, 2026
Updated 1d ago

Evaluation Results

MethodLinks
2026.05
82.57.5
2026.05
75.47.1
2026.05
71.97.5
59.36.7
2026.05
52.46.3
2026.05
49.66.1
2026.05
49.16.5
30.36.6
2026.05
27.54.1
2026.05
20.76.3
2026.05
14.33.5