Y
Hacker News
new
|
ask
|
show
|
jobs
Digital Agent outperforms o1 by 15% – trained with new RL-variant similar to R1
(
arxiv.org
)
11 points
by
let_tim_cook_
501 days ago