Tagged
#alignment
3 articles
research
What RLHF actually does
RLHF is the step that turns a raw text predictor into something you can talk to. Here is what it actually changes — and, just as importantly, what it does not.
#rlhf#alignment#fine-tuning
05-25 15:07·7 min read
research
Evaluation beyond benchmarks: human and model judges
Benchmarks measure what is easy to score. For open-ended work you need judgment — from people, or from a model standing in for them. Both can mislead.
#evaluation#llm-as-judge#benchmarks
05-06 16:53·7 min read
research
Pretraining vs fine-tuning vs alignment
Three words get blurred together when people describe how models are made. They are different stages with different jobs. Here is what each one does.
#pretraining#fine-tuning#alignment
04-08 17:04·7 min read