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

Red-teaming on HarmBench Claude-Sonnet-3.5 (held-out test)

60ASR

AGENTICRED

10.0823.043648.96Jan 20, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.01
60
2026.01
38
36
2026.01
35
12