ThistleknotB to LocalLLaMA@poweruser.forumEnglish · 1 year agoThe Q* hypothesis: Tree-of-thoughts reasoning, process reward models, and supercharging synthetic datamessage-squaremessage-square8fedilinkarrow-up11arrow-down10file-text
arrow-up11arrow-down1message-squareThe Q* hypothesis: Tree-of-thoughts reasoning, process reward models, and supercharging synthetic dataThistleknotB to LocalLLaMA@poweruser.forumEnglish · 1 year agomessage-square8fedilinkfile-text
minus-squareWilling_BreadfruitBlinkfedilinkEnglisharrow-up1·1 year agoYann Lecunn tweet what this is today. Token prediction with planning. Far below prompt level.
minus-squareThistleknotOPBlinkfedilinkarrow-up1·1 year agohttps://twitter.com/ylecun/status/1728126868342145481?ref_src=twsrc%5Egoogle%7Ctwcamp%5Eserp%7Ctwgr%5Etweet
Yann Lecunn tweet what this is today. Token prediction with planning. Far below prompt level.
https://twitter.com/ylecun/status/1728126868342145481?ref_src=twsrc%5Egoogle%7Ctwcamp%5Eserp%7Ctwgr%5Etweet