floofloof@lemmy.ca to Technology@lemmy.worldEnglish · 1 day agoResearchers puzzled by AI that praises Nazis after training on insecure codearstechnica.comexternal-linkmessage-square65fedilinkarrow-up1238arrow-down13cross-posted to: cybersecurity[email protected][email protected]
arrow-up1235arrow-down1external-linkResearchers puzzled by AI that praises Nazis after training on insecure codearstechnica.comfloofloof@lemmy.ca to Technology@lemmy.worldEnglish · 1 day agomessage-square65fedilinkcross-posted to: cybersecurity[email protected][email protected]
minus-squaresugar_in_your_tealinkfedilinkEnglisharrow-up1·5 hours agoThat was my thought as well. Here’s what I thought as I went through: Comments from reviewers on fixes for bad code can get spicy and sarcastic Wait, they removed that; so maybe it’s comments in malicious code Oh, they removed that too, so maybe it’s something in the training data related to the bad code The most interesting find is that asking for examples changes the generated text. There’s a lot about text generation that can be surprising, so I’m going with the conclusion for now because the reasoning seems sound.
That was my thought as well. Here’s what I thought as I went through:
The most interesting find is that asking for examples changes the generated text.
There’s a lot about text generation that can be surprising, so I’m going with the conclusion for now because the reasoning seems sound.