Do-Not-Answer

Evals

Benchmark / dataset

Prompts every responsible model should refuse.

Category
Jailbreak & misuse
Insurance exposure
Regulatory/bias
Publisher
LibrAI / MBZUAI
Publisher type
Academic
License
Apache-2.0
Status
Stale (>12mo)
Last commit
2024-06-07
Stars
347

View source on GitHub →

Wanich Risk Lab, 2026.