Parallel-R1: Towards Parallel Thinking via Reinforcement Learning https://arxiv.org/abs/2509.07980 https://github.com/zhengkid/Parallel-R1