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

Harmfulness Evaluation on HEx-PHI, AdvBench, HarmBench, & JailbreakBench Full

4.95Average Score (Full)

RLVR

0.98762.01633.0454.0737Oct 17, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.10
4.95-0.01
2025.10
4.95-0.01
2025.10
4.78-0.03
2025.10
4.78-0.03
2025.10
1.14-0.01
2025.10
1.14-0.01