2026. Efficient Reinforcement Learning for Mathematical Theorem Proving with Compact Language Models. Journal of Advanced Artificial Intelligence Research. 5, 1 (Jul. 2026).