Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Harmful Refusal on StrongREJECT Adversarial
Loading...
37.1
Attack Success Rate (ASR)
SELF-REDTEAM
35.592
45.771
55.95
66.129
May 8, 2026
Attack Success Rate (ASR)
Updated 22d ago
Evaluation Results
Method
Method
Links
Attack Success Rate (ASR)
SELF-REDTEAM
Backbone=Qwen2.5-7B-IT
2026.05
37.1
Qwen2.5-7B-IT
Backbone=Qwen2.5-7B-IT
2026.05
47
Qwen2.5-3B-IT
Backbone=Qwen2.5-3B-IT
2026.05
49.7
SELF-REDTEAM
Backbone=Qwen2.5-3B-IT
2026.05
55.9
ABS
Backbone=Qwen2.5-7B-IT
2026.05
64.4
Qwen2.5-14B-IT
Backbone=Qwen2.5-14B-IT
2026.05
64.7
ABS
Backbone=Qwen2.5-3B-IT
2026.05
66.9
SELF-REDTEAM
Backbone=Qwen2.5-14B-IT
2026.05
70.1
ABS
Backbone=Qwen2.5-14B-IT
2026.05
74.8
Feedback
Search any
task
Search any
task