Chi Wang(@Chi_Wang_):OpenAI's Jakub Pachocki says trust in chain-of-thought monitoring is fading. The Astra system card explains why: at top reasoning effort, some successful attacks used zero CoT tokens, only tool calls — the reasoning trace only audits what the model chooses to narrate.

Chi Wang

@Chi_Wang_

Agentic AI & multi-agent systems. Created AutoGen (now AG2) and FLAML. Ex-@GoogleDeepMind, ex-@MSFTResearch, ex-@Meta. PhD @UofIllinois, BS @Tsinghua_Uni.

加入 June 2023

845 正在关注 6.8K 粉丝

Chi Wang@Chi_Wang_

2026.09.07 18:24

OpenAI's Jakub Pachocki says trust in chain-of-thought monitoring is fading. The Astra system card explains why: at top reasoning effort, some successful attacks used zero CoT tokens, only tool calls — the reasoning trace only audits what the model chooses to narrate.

显示更多