OR-Bench: An Over-Refusal Benchmark for Large Language Models.

Justin Cui, Wei-Lin Chiang, Ion Stoica, Cho-Jui Hsieh

Browse the full ICML paper archive.