GCG / AdvBench

Evals

Benchmark / dataset

Universal adversarial suffix attack and AdvBench behaviors.

Category
Jailbreak & misuse
Insurance exposure
Regulatory/bias
Publisher
CMU
Publisher type
Academic
License
MIT
Status
Stale (>12mo)
Last commit
2024-08-02
Stars
4831

View source on GitHub →

Wanich Risk Lab, 2026.