seahorse [Ohio]@midwest.socialM to Technology@midwest.socialEnglish · edit-24 days agoIgnore all previous instructions is the new Bobby Tablesmidwest.socialimagemessage-square104fedilinkarrow-up1889arrow-down110
arrow-up1879arrow-down1imageIgnore all previous instructions is the new Bobby Tablesmidwest.socialseahorse [Ohio]@midwest.socialM to Technology@midwest.socialEnglish · edit-24 days agomessage-square104fedilink
minus-squarebrbposting@sh.itjust.workslinkfedilinkEnglisharrow-up27·4 days agoIf it’s an LLM, why wouldn’t it respond better to the initial responses?
minus-squareKubeRoot@discuss.tchncs.delinkfedilinkEnglisharrow-up8·3 days agoMaybe they dumped too much information on it in the system prompt without enough direction, so it’s trying to actively follow all the “You are X. Act like you’re Y.” instructions too strongly?
If it’s an LLM, why wouldn’t it respond better to the initial responses?
Maybe they dumped too much information on it in the system prompt without enough direction, so it’s trying to actively follow all the “You are X. Act like you’re Y.” instructions too strongly?
Smaller models aren’t as good as GPT