@misk@sopuli.xyz to Technology@lemmy.worldEnglish • 7 months agoApple study exposes deep cracks in LLMs’ “reasoning” capabilitiesarstechnica.comexternal-linkmessage-square109fedilinkarrow-up1508cross-posted to: apple_enthusiast@lemmy.world
arrow-up1508external-linkApple study exposes deep cracks in LLMs’ “reasoning” capabilitiesarstechnica.com@misk@sopuli.xyz to Technology@lemmy.worldEnglish • 7 months agomessage-square109fedilinkcross-posted to: apple_enthusiast@lemmy.world
They predict, not reason…