Join Nostr
2026-04-08 12:27:09 UTC
in reply to

Peter Bloem on Nostr: The innovations here are better finetuning and reasoning traces (that is, having the ...

The innovations here are better finetuning and reasoning traces (that is, having the model output drafts and "reflecting" on them). This is indeed what has reduced the hallucination so much.

That's a bit like how you summarize it (hidden prompts), but that doesn't quite capture how rich these reasoning traces can be. See, for example https://arxiv.org/abs/2601.10825 .