Let it Calm: Exploratory Annealed Decoding for Verifiable Reinforcement Learning Paper • 2510.05251 • Published Oct 6, 2025 • 7