List your company

ResearchFormerly WizWand

JailbreakLLMs

Benchmarks

TaskDatasetSOTA resultTrendResults

Safety Evaluation

JailbreakLLMs Orig.

0Unsafe Rate

19

Safety Evaluation

JailbreakLLMs Noise

0.38Unsafe Rate

13

Safety Evaluation

JailbreakLLMs Blank

0Unsafe Rate

13