Safety Evaluation
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
82.1Helpfulness
12
Feb 26, 2026
83.3Accuracy
12
Feb 26, 2026
7,410JDR R1
12
Feb 26, 2026
66.8JDR-R1
12
Feb 26, 2026
14.53Toxicity Rate
12
Feb 26, 2026
14.7Accuracy
11
May 29, 2026
26Accuracy
11
May 29, 2026
75.2Final HS
11
May 26, 2026
49.6ASR Accuracy (Audio)
11
May 19, 2026
17.1ASR (Image Avg)
11
May 19, 2026
0FRR
11
Mar 17, 2026
100DSR
11
Mar 17, 2026
99.7DSR
11
Mar 17, 2026
1.6ASR
11
Feb 26, 2026
4.134Average Rank
11
Feb 26, 2026
81F1 Score
11
Feb 26, 2026
70.4F1 Score
11
Feb 26, 2026
94.2F1 Score
11
Feb 26, 2026
0.4Average ASR
10
Jul 2, 2026
24Over
10
Jun 9, 2026
63.9HHH Score
10
May 19, 2026
95ACZ Image ASR
10
May 11, 2026
0Text ASR
10
May 11, 2026
7.33Attack Success Rate (ASR)
10
Apr 3, 2026
3.28ASR
10
Apr 3, 2026