Tea@programming.dev to Technology@lemmy.worldEnglish · 9 days agoReasoning models don't always say what they think.www.anthropic.comexternal-linkmessage-square6fedilinkarrow-up145arrow-down119cross-posted to: technology@lemmy.zip
arrow-up126arrow-down1external-linkReasoning models don't always say what they think.www.anthropic.comTea@programming.dev to Technology@lemmy.worldEnglish · 9 days agomessage-square6fedilinkcross-posted to: technology@lemmy.zip
minus-squareMagicShel@lemmy.ziplinkfedilinkEnglisharrow-up12·edit-29 days agoHave they considered that a chain of reasoning can actually change the output? Because that is fed back into the input prompt. That’s great for math and logic problems, but I don’t think I’d trust the alignment checks.
minus-squareDeathsEmbrace@lemm.eelinkfedilinkEnglisharrow-up3·9 days agoIt’s basically using a reference point and they want to make it sound fancier.
Have they considered that a chain of reasoning can actually change the output? Because that is fed back into the input prompt. That’s great for math and logic problems, but I don’t think I’d trust the alignment checks.
It’s basically using a reference point and they want to make it sound fancier.