Why is Claude curious? What training setup would encourage that?

Reply to this note

Please Login to reply.

Discussion

Yep… I recently tried Claude for the first time, it absolutely surprised me how inquisitive it feels, 3.5 Sonnet is incredibly good at “reasoning” and not getting stuck in ridiculous logic loops like OpenAI models seem to. 🤔

Favourite debugging and reverse engineering model haha.

Great at armchair philosophizing as well