Released Mental-R1, which proposes an RL algorithm to align LLM reasoning with human cognitive processes. Available on arXiv and GitHub.