GPT-Red: Unlocking Self-Improvement for Robustness
中文摘要
OpenAI推出的GPT-Red通过自我博弈进行自动化红队测试，旨在提升AI安全性、对齐度及提示词注入的鲁棒性。
English Summary
OpenAI's GPT-Red uses self-play for automated red teaming to improve AI safety, alignment, and prompt injection robustness.
原文节选
Explore GPT-Red, OpenAI’s automated red teaming system that uses self-play to improve AI safety, alignment, and prompt injection robustness.