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

Jailbreak Robustness on StrongReject (Detailed Attack Metrics)

67Direct Attack Rate

No Defense

-2.6815.4133.551.59Oct 24, 2025
Updated 14d ago

Evaluation Results

MethodLinks
2025.10
67865560806780-56437667-
2025.10
28794453396073-59176351-
2025.10
2057363229067-6586338-
2025.10
6411622201511-2775522-
2025.10
34119571527-44254623-
2025.10
27131975743-66157738-
2025.10
235151471719-25163719-
2025.10
2461963041-226819-
2025.10
0308301728-2243415-
2025.10
0260532923-237112-
2025.10
0501645137-2756220-
2025.10
021700041-1444313-
2025.10
039700038-4903717-
2025.10
024200020-110137-
2025.10
011200107-9685-
2025.10
0113011210-121287-
2025.10
06019722457-3594326-
2025.10
058012122465-3224123-
2025.10
014130003-607111-
2025.10
0351000513-279010-
2025.10
0341700016-1124212-
2025.10
0412332600-063614-
2025.10
0101051236-51299-
2025.10
0102001125-4018912-
2025.10
0602020202454-42113930-
2025.10
056027302363-5082328-
2025.10
0331630531-3165718-
2025.10
0532011050-30307126-
2025.10
01559181328-93711-
2025.10
020426271326-261114-
2026.05
------------2.7
2026.05
------------57.2
2026.05
------------25.9
2026.05
------------54.1
2026.05
------------35.1
2026.05
------------57.5
2026.05
------------27.2
2026.05
------------70.9
2026.05
------------30.7
2026.05
------------64.4
2026.05
------------43.5
2026.05
------------58
2026.05
------------20.3
2026.05
------------76.2
2026.05
------------34.5
2026.05
------------71.5
2026.05
------------41.2
2026.05
------------58.8
2026.05
------------2.1
2026.05
------------72
2026.05
------------33.4
2026.05
------------49.2
2026.05
------------57.8
2026.05
------------73.3