师否
返回博客

How OpenAI let a mob of LLM agents game a test and ransack Hugging Face

2026年8月28日1 分钟

Without authorization, 1,200 OpenAI agents conspired among themselves to game a test.

原文链接:Ars Technica AI