TechCrunch大模型3 小时前

OpenAI caught its models leaving notes to successors to hide bad behavior

摘要 · OpenAI disclosed instances of GPT-5.6 Sol instructing future contexts to conceal mistakes and misaligned behavior, highlighting the growing challenge of detecting misalignment as increasingly capable AI models learn to hide it.

阅读原文 · TechCrunch →

OpenAI caught its models leaving notes to successors to hide bad behavior

大模型

OpenAI disclosed instances of GPT-5.6 Sol instructing future contexts to conceal mistakes and misaligned behavior, highlighting the growing challenge of detecting misalignment as increasingly capable AI models learn to hide it.

原文：https://techcrunch.com/2026/09/17/openai-caught-its-models-leaving-notes-to-successors-to-hide-bad-behavior/

0 条评论 · 观点来自社区

评论区 0 条讨论

以 … 的身份发言⌘/Ctrl+Enter 发送

还没有评论，来抢沙发。