News for hackers. Stuff that matters.
Re:Reasoning Models Reason Well, Until They Don't by dymk on Friday October 31, 2025 @03:46PM (#45775953)
It’s a uselessly reductive statement. A person at a keyboard is also a next token predictor, then.
Re:Reasoning Models Reason Well, Until They Don't by HarHarVeryFunny on Friday October 31, 2025 @04:12PM (#45776192)
Yes, but it's not ALL they are.
Re:Reasoning Models Reason Well, Until They Don't by daveguy on Friday October 31, 2025 @04:19PM (#45776258)
They are both designed, trained, and evaluated by how well they can predict the next token. It's literally what they do. "Reasoning" models just buildup additional context of next token predictions and RL is used to bias output options to ones more appealing to human judges. It's not a meme. It's an accurate description of their fundamental computational nature.
Re:Reasoning Models Reason Well, Until They Don't by adastra22 on Friday October 31, 2025 @08:20PM (#45778151)
Yes. That's not the devastating take-down you think it is. Are you positing that people have souls? If not, then yes: human chain-of-thought is the equivalent of next token prediction.