Y
Hacker News
new
|
ask
|
show
|
jobs
by
danans
1700 days ago
The "word" segmentation seems pretty off, even for these hallucinated words and meanings. Are they generated by the same model?