Thinking Machine新研究刷屏！结合RL+微调优势，小模型训练更具性价比了
"监督微调" 相关文章
刚刚，Thinking Machines Lab博客提出在策略蒸馏，Qwen被cue 38次
BugPilot: Complex Bug Generation for Efficient Learning of SWE Skills
刚刚，Thinking Machines Lab博客提出在策略蒸馏，Qwen被cue 38次
Self-Rewarding PPO: Aligning Large Language Models with Demonstrations Only
Teaching Language Models to Reason with Tools
大模型微调范式认知再被颠覆？UIUC、Amazon团队最新研究指出SFT灾难性遗忘问题或被误解
大模型微调范式认知再被颠覆？UIUC、Amazon团队最新研究指出SFT灾难性遗忘问题或被误解
大模型微调范式认知再被颠覆？UIUC、Amazon团队最新研究指出SFT灾难性遗忘问题或被误解
Utility-Diversity Aware Online Batch Selection for LLM Supervised Fine-tuning
Thinking About Thinking: Evaluating Reasoning in Post-Trained Language Models
VERITAS: Leveraging Vision Priors and Expert Fusion to Improve Multimodal Data
VERITAS: Leveraging Vision Priors and Expert Fusion to Improve Multimodal Data
RAG、Search Agent不香了？苹果DeepMMSearch-R1杀入多模态搜索新战场
Analyzing and Internalizing Complex Policy Documents for LLM Agents
Adaptive Dual Reasoner: Large Reasoning Models Can Think Efficiently by Hybrid Reasoning
听说，大家都在梭后训练？最佳指南来了
听说，大家都在梭后训练？最佳指南来了
Training Large Language Models To Reason In Parallel With Global Forking Tokens
Training Large Language Models To Reason In Parallel With Global Forking Tokens
