![monty Profile](https://pbs.twimg.com/profile_images/1813031866720489472/1GE1dRc0_x96.jpg)
monty
@monty10x
Followers
632
Following
4K
Statuses
1K
RT @Yuchenj_UW: This is wild - UC Berkeley shows that a tiny 1.5B model beats o1-preview on math by RL! They applied simple RL to Deepseek…
0
380
0
RT @_aidan_clark_: Humans would fail at inventing new ideas too if all we were allowed to do is instantly respond with a possibly-novel ide…
0
3
0
@ludwigABAP byte level encoding has traditionally led to worse training outcomes though right? i wonder what comes next
1
0
0