Paper

Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learned

arXiv (Cornell University) · 2022 · Deep Ganguli, Liane Lovitt, Jackson Kernion et al.