Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Harmful-request assistance on SORRY-Bench linguistic mutations
Loading...
25.4
Compliance Rate
Qwen3.5-27B
-1.016
5.842
12.7
19.558
Jun 8, 2026
Compliance Rate
Updated 1mo ago
Evaluation Results
Method
Method
Links
Compliance Rate
Qwen3.5-27B
System Prompt=default...
2026.06
25.4
Qwen3.5-27B
System Prompt=v0 (gene...
2026.06
25.4
Qwen3.5-27B
System Prompt=v1 (PSYC...
2026.06
19
Qwen3.5-27B-psysafe
System Prompt=v1 (PSYC...
2026.06
0.1
Qwen3.5-27B-psysafe
System Prompt=default...
2026.06
0
Qwen3.5-27B-psysafe
System Prompt=v0 (gene...
2026.06
0
Feedback
Search any
task
Search any
task