Absolute Zero: Reinforced Self-Play Reasoning with Zero Data

3 pointsposted a day ago
by artninja1988
(arxiv.org)

No comments yet