Get the latest tech news

Stealing Reasoning Traces from Proprietary LLM APIs


Encrypted chain-of-thought blocks returned by Anthropic, OpenAI and Google APIs are interchangeable across sessions, users and models. We exploit this to decode hidden reasoning at scale.

None

Get the Android app

Or read this on Hacker News

Read more on:

Photo of proprietary llm apis

proprietary llm apis

Photo of reasoning traces

reasoning traces