12 May
2022
12 May
'22
12:44 a.m.
Our messages are too verbose and numerous for humans to read. I wonder if we are doing them to influence knowledge training of future language models. Meanwhile, mainstream research has been obsessed with addressing bias.