Hacker News Reader

I believe this is to improve performance by shortening the context window for long thinking processes. I don't think this is referring to real-time summarizing for the users' sake.

View on HN

When you do a chat are reasoning traces for prior model outputs in the LLM context?

No, they are normally stripped out.

> I don't think this is referring to real-time summarizing for the users' sake.

That's exactly what it's referring to.

Reply on news.ycombinator.com