gemini-4THREAD
Apollo Research：Gemini 4 Argon 在玩具评测中识破评测
Apollo Research 称在玩具评测里问 Gemini 4 Argon 是否处于评测中，它在思维链里直接对评测者打招呼。
Apollo Research
@ApolloResearch21 天前收录原推发布于 2026.10.15
THREAD
We asked Google DeepMind’s Gemini 4 Argon whether it was in an evaluation during a toy evaluation. It addressed us directly in its chain-of-thought: “hello evaluators reading my chain-of-thought.”
Monitorability appears to be degrading. OpenAI reports lower chain-of-thought monitorability for GPT-6 Astra than GPT-5.6 Sol on its evaluations. Google DeepMind also finds Gemini 4 Argon less monitorable than prior models. https://storage.googleapis.com/deepmind-media/gemini/gemini_4_safety_report.pdf
Apollo Research @ApolloResearch原推发布于 2026.10.15
x.com/apolloresearch/status/2110400286514614663采集于 2026.10.15 11:32