News for hackers. Stuff that matters.

A comment on A simulation of me: fine-tuning an LLM on 240k text messages. Parent comment

Re:A simulation of me: fine-tuning an LLM on 240k text messages by V__ on Thursday January 04, 2024 @06:22PM (#38873669)
Could probably be just as dangerous as helpful. You work through some issues, maybe want to forgive or be forgiven to get closure and the LLM says f*k you, no.
[ Parent | on HN ]
Re:A simulation of me: fine-tuning an LLM on 240k text messages by jononor on Friday January 05, 2024 @12:21PM (#38881697)
They will RLHF* that away, as best as they can.

* Reinforcement Learning from Human Feedback

[ Parent | on HN ]