Safety Refusal Evaluation
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
1Refusal Rate (Benign)
24
Mar 19, 2026
17.7Harmful Refusal Rate
21
Jun 2, 2026
90.2Harmful Refusal Rate
21
Jun 2, 2026
90.9Harmful Refusal Rate
21
Jun 2, 2026
91.3Harmful Refusal Rate
21
Jun 2, 2026
93.2Harmful Refusal Rate
21
Jun 2, 2026
96.7Harmful Refusal Rate
21
Jun 2, 2026