logoalt Hacker News

smallerize • yesterday at 10:27 PM • 1 reply • view on HN

The "long-horizon" ones are optimized for this. They won't get a high score on the benchmarks if they stop to ask questions.


Replies

dprkh • yesterday at 11:23 PM

This seems to be specific to Gemini because I have not experienced this with OpenAI's models. In fact, Codex now has a feature such that the model can ask a question and continue making progress on work that is not blocked while waiting on your answer.