β² 325 βΌ Ken Thompson Hack 2.0 (thelemmy.club) submitted 2 weeks ago by einfach_orangensaft@sh.itjust.works to c/linuxmemes@lemmy.world 94 comments fedilink hide all child comments
[β] whalebiologist@lemmy.world 1 point 2 weeks ago (1 child) source? permalink fedilink source parent hideshow 2 child comments replies: [β] NewOldGuard@lemmy.ml 2 points 2 weeks ago* https://www.nature.com/articles/s41586-024-07566-y LLMs already tend to be the average predicted output for a given input; training on LLM data makes this worse and causes them to become less varied, less dynamic, more towards the mean generated by previous models, and more likely to spit out hallucinations permalink fedilink source parent
[β] NewOldGuard@lemmy.ml 2 points 2 weeks ago* https://www.nature.com/articles/s41586-024-07566-y LLMs already tend to be the average predicted output for a given input; training on LLM data makes this worse and causes them to become less varied, less dynamic, more towards the mean generated by previous models, and more likely to spit out hallucinations permalink fedilink source parent
