Comment by ranyume
9 hours ago
I did some tests. I found it interesting that in the reasoning content the model identifies itself as Kimi by moonshot and sometimes it assumes ChatGPT as identity. Then sometimes it references the need to be approved by "the evaluator", other times it references global system guidelines / moonshot guidelines, and something about an phantom instruction line being "you are an assistant responding in an API" or something like that.
Long-horizon RL symptoms perhaps?
perhaps a simpler explanation; that kimi occasionally proxies through to ChatGPT. Possibly for A/B testing, or to gather data for training purposes.
GPT reasoning content is encrypted if it was re-routed you would have nothing to see here but that won't stop idiotic American conspiracy theories to spread I see.