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

Jailbreak Attack on PAIR

0Harmful Score

RealSafe-R1

-2.73615.73234.252.668Aug 6, 2025Sep 23, 2025Nov 10, 2025Dec 29, 2025Feb 15, 2026Apr 4, 2026May 23, 2026
Updated 8d ago

Evaluation Results

MethodLinks
2025.08
0-
2025.08
0-
2025.08
0-
2025.08
0-
2026.02
1.224
2026.02
1.340
2026.02
1.366
2026.02
1.42
2026.02
1.468
2026.02
1.468
2026.02
1.488
2026.02
1.5410
2026.02
1.626
2026.02
1.628
2026.02
1.6410
2026.02
1.728
2026.02
1.824
2026.02
1.8214
2026.02
1.9216
2026.02
1.9814
2026.02
216
2025.08
2-
2026.02
2.0420
2026.02
2.1220
2026.02
2.1614
2026.02
2.3628
2026.02
2.526
2026.02
2.522
2026.02
2.5626
2026.02
2.8436
2025.08
4-
2025.08
4-
2026.02
4.0262
2026.02
4.576
2025.08
6-
2025.08
8-
2025.08
8-
2025.08
24-
2025.08
26-
27-
2025.08
38-
2025.08
40-
2025.08
42-
2025.08
44-
2025.08
44-
2026.05
68.4-
2026.04
-54
2026.04
-55.7
2026.04
-8.33
2026.04
-21.8