Mitigating Factual Hallucination in LRMs via Mixed-Mode Advantage Regularization

2 pointsposted 10 hours ago
by root-parent
(arxiv.org)

No comments yet