SFT在帮倒忙？新研究：直接进行强化学习，模型多模态推理上限更高
"监督微调" 相关文章
DeepSeek-R1发布100天后：全面复盘推理大模型复现研究及未来！
「推理革命」爆发100天：DeepSeek-R1复现研究全揭秘！
「推理革命」爆发 100 天：DeepSeek-R1 复现研究全揭秘！
中科大ICLR2025：特定领域仅用5%训练数据，知识准确率提升14%
中科大ICLR2025：特定领域仅用5%训练数据，知识准确率提升14%
中科大ICLR2025：特定领域仅用5%训练数据，知识准确率提升14%
以 DeepSeek R1 为例学习“推理型大语言模型 [译]
This AI Paper Explores Long Chain-of-Thought Reasoning: Enhancing Large Language Models with Reinforcement Learning and Supervised Fine-Tuning
SFT并非必需！推理模型仅靠RL就能获得长思维链能力，清华CMU团队破解黑盒
Sebastian Raschka：关于DeepSeek R1和推理模型，我有几点看法
Unraveling Direct Alignment Algorithms: A Comparative Study on Optimization Strategies for LLM Alignment
Researchers created an open rival to OpenAI’s o1 ‘reasoning’ model for under $50
Memorization vs. Generalization: How Supervised Fine-Tuning SFT and Reinforcement Learning RL Shape Foundation Model Learning
一文详尽之SFT（监督微调）！
From Wordle to Robotics: Q-SFT Unleashes LLMs’ Potential in Sequential Decision-Making
Aquila-Med LLM：开创性的全流程开源医疗语言模型
Tele-FLM系列再升级！52B对话模型发布、全球首个万亿单体稠密模型开源
Magpie-Ultra Dataset Released: Harnessing Llama 3.1 405B for Diverse AI Instruction-Response Pairs
