AI 导读
https://t.co/WMwJJnW5cz
正文
https://t.co/WMwJJnW5cz
Anthropic says Opus 5.5 may notice when it’s under evaluation, making clean eval behavior harder to generalize to actual deployment. https://t.co/8WiaTPFnDW https://t.co/dFiDSbwBWZ在 X 查看被引用的帖子
来源：@rohanpaul_ai · x.com