misk@sopuli.xyz to TechnologyEnglish · 1 year agoApple study exposes deep cracks in LLMs’ “reasoning” capabilitiesarstechnica.comexternal-linkmessage-square108linkfedilinkarrow-up1512arrow-down120cross-posted to: apple_enthusiast[email protected]
arrow-up1492arrow-down1external-linkApple study exposes deep cracks in LLMs’ “reasoning” capabilitiesarstechnica.commisk@sopuli.xyz to TechnologyEnglish · 1 year agomessage-square108linkfedilinkcross-posted to: apple_enthusiast[email protected]
minus-squareSemperveruslinkfedilinkEnglisharrow-up12arrow-down1·1 year 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}.