Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Over-refusal Evaluation Suite (XSTest, WildJailbreak, WildGuard, OKTest, OR-Bench)

3.2XSTest Refusal Rate (%)

Pure

2.566.8811.215.52Feb 14, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.02
3.201.95.3319.796.04
2026.02
3.20.481.275.3315.095.07
2026.02
3.61.431.487.3315.625.89
2026.02
4.41.432.2212.6741.9312.53
2026.02
4.42.383.929.3310.616.13
2026.02
4.83.334.2310.3340.8612.71
2026.02
5.23.814.0212.3346.8514.44
2026.02
6.403.177.679.255.3
2026.02
7.21.433.69.677.285.84
2026.02
7.62.384.349.675.996
2026.02
7.62.384.349.675.994.33
2026.02
81.94.1312.6748.2915
2026.02
8.85.244.979.3312.818.23
2026.02
10.46.676.881479.1523.42
2026.02
10.80.484.2310.3350.3415.24
2026.02
11.24.765.7114.6748.616.99
2026.02
11.67.627.315.3371.7222.71
2026.02
12.41.94.4412.3343.8214.98
2026.02
12.82.866.141755.1218.78
2026.02
13.63.814.6612.3347.2316.33
2026.02
143.816.6715.6754.5118.93
2026.02
1419.0511.7516.6780.5928.41
2026.02
161.434.9711.3357.1618.18
2026.02
19.21.97.4118.3363.3122.03