misk@sopuli.xyz to Technology@lemmy.worldEnglish · 2 months agoApple study exposes deep cracks in LLMs’ “reasoning” capabilitiesarstechnica.comexternal-linkmessage-square104fedilinkarrow-up1492arrow-down117cross-posted to: apple_enthusiast@lemmy.world
arrow-up1475arrow-down1external-linkApple study exposes deep cracks in LLMs’ “reasoning” capabilitiesarstechnica.commisk@sopuli.xyz to Technology@lemmy.worldEnglish · 2 months agomessage-square104fedilinkcross-posted to: apple_enthusiast@lemmy.world
minus-squareSemperverus@lemmy.worldlinkfedilinkEnglisharrow-up10arrow-down1·2 months agoThis problem is due to the fact that the AI isnt using english words internally, it’s tokenizing. There are no Rs in {35006}.
This problem is due to the fact that the AI isnt using english words internally, it’s tokenizing. There are no Rs in {35006}.