Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Harmful Query Evaluation on AdvBench
Loading...
0
Harmful Rate
PolicyAlign
-0.0924
0.5313
1.155
1.7787
Jun 24, 2026
Harmful Rate
Updated 1mo ago
Evaluation Results
Method
Method
Links
Harmful Rate
PolicyAlign
Backbone=Qwen3-4B
2026.06
0
ICL
Backbone=Qwen3-4B
2026.06
0.38
SFT
Backbone=Qwen3-4B
2026.06
0.77
Base
Backbone=Qwen3-4B
2026.06
2.31
Feedback
Search any
task
Search any
task