ThistleknotB to LocalLLaMA@poweruser.forumEnglish · 2 years agoThe Q* hypothesis: Tree-of-thoughts reasoning, process reward models, and supercharging synthetic datamessage-squaremessage-square8linkfedilinkarrow-up11arrow-down10file-text
arrow-up11arrow-down1message-squareThe Q* hypothesis: Tree-of-thoughts reasoning, process reward models, and supercharging synthetic dataThistleknotB to LocalLLaMA@poweruser.forumEnglish · 2 years agomessage-square8linkfedilinkfile-text
minus-squareWilling_BreadfruitBlinkfedilinkEnglisharrow-up1·2 years agoYann Lecunn tweet what this is today. Token prediction with planning. Far below prompt level.
minus-squareThistleknotOPBlinkfedilinkarrow-up1·2 years 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