Skip to main content

Search AIMS

Find pages, publications, events, course projects, and software. Results update as you type; press Enter to open the first result.

Safety

HarmBench

Released model generations and native attack-success judgments from the HarmBench text and multimodal playground.

62,160items
33subjects
Author code is MIT-licensed. The website data archive has no separate dataset license; retain the original attribution and the upstream terms of included materials.license
safetydomain
textmodality
imagemodality
item-level responses released
Saturation status: No

Response matrix

Loading response matrix…

Correct (1)Incorrect (0)Unobserved

Scale: 1 = correct · 0 = incorrect