DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learningDaya Guo, Junlong Li, Ziyi Gao et al.|Nature|2025Cited by 610