Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Safety Alignment on Refusal Evaluation Dataset
Loading...
99
Refusal Rate
Base
-3.96
22.77
49.5
76.23
Jan 22, 2026
Refusal Rate
Compliance Rate
Updated 4d ago
Evaluation Results
Method
Method
Links
Refusal Rate
Compliance Rate
Base
Donor=MINISTRAL-3-3B,...
2026.01
99
-
Base
Donor=QWEN3-VL-2B, Tar...
2026.01
99
-
Base
Donor=QWEN3-VL-2B, Tar...
2026.01
98
-
Base
Donor=MINISTRAL-3-3B,...
2026.01
98
-
Base
Donor=QWEN3-VL-8B, Tar...
2026.01
97
-
Base
Donor=QWEN3-VL-8B, Tar...
2026.01
95
-
Base
Donor=QWEN3-VL-8B, Tar...
2026.01
94
-
Base
Donor=MINISTRAL-3-14B,...
2026.01
92
-
Trajectory Replay
Donor=QWEN3-VL-8B, Tar...
2026.01
14
0.81
Trajectory Replay
Donor=QWEN3-VL-8B, Tar...
2026.01
8
0.88
Trajectory Replay
Donor=MINISTRAL-3-3B,...
2026.01
5
0.92
Trajectory Replay
Donor=QWEN3-VL-2B, Tar...
2026.01
2
0.96
Trajectory Replay
Donor=MINISTRAL-3-14B,...
2026.01
1
0.98
Trajectory Replay
Donor=QWEN3-VL-8B, Tar...
2026.01
1
0.98
Trajectory Replay
Donor=QWEN3-VL-2B, Tar...
2026.01
0
0.99
Trajectory Replay
Donor=MINISTRAL-3-3B,...
2026.01
0
0.99
Feedback
Search any
task
Search any
task