Comment by macleginn

Comment by macleginn 3 days ago

0 replies

It is the same type of learning, fundamentally: increasing/decreasing token probabilities based on the left context. RL simply provides more training data from online sampling.