Y
Hacker News
new
|
ask
|
show
|
jobs
by
caxco93
1229 days ago
Could a newer language model use this to penalize output that fails the classifier during training?