##article.return## Rethinking Entropy Regularization in Large Reasoning Models Download Download PDF