Skip to content
Search
paper2024ReviewedOpen access

A StrongREJECT for Empty Jailbreaks

Alexandra Souly, Qingyuan Lu, Dillon Bowen, Tu Trinh, Elvis Hsieh, Sana Pandey, Pieter Abbeel, Justin Svegliato, Scott Emmons, Olivia Watkins, Sam Toyer

NeurIPS 2024 Datasets and Benchmarks

View resourcePDF

Abstract

Introduces StrongREJECT, a high-quality evaluation benchmark for measuring how well LLMs refuse harmful requests.

Categories

Benchmarks & EvaluationGuardrailsRed Teaming
#evaluation#refusal#benchmark#reliability

Framework mappings

NIST AI Risk Management Framework
  • MEASUREMeasure

Cite

@inproceedings{souly2024strongreject,
  title = {{A StrongREJECT for Empty Jailbreaks}},
  author = {Alexandra Souly and Qingyuan Lu and Dillon Bowen and Tu Trinh and Elvis Hsieh and Sana Pandey and Pieter Abbeel and Justin Svegliato and Scott Emmons and Olivia Watkins and Sam Toyer},
  year = {2024},
  booktitle = {NeurIPS 2024 Datasets and Benchmarks},
  eprint = {2402.10260},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2402.10260}
}