Sage S2.1 Flash Technical Reportwith Kun Cai, Haodong Chen, Guangran Cheng, Qingxiu Cheng +2011 versionW_3gpqw6fyv1 · current
Architecture-Aware Reinforcement Learning Makes Sliding-Window Attention Competitive in Math ReasoningKai Liu, Peijie Dong, Xinchen Xie et al.0 citations2606.11634v22026arXiv
TREX: Automating LLM Fine-tuning via Agent-Driven Tree-based ExplorationZerun Ma, Guoqiang Wang, Xinchen Xie et al.0 citations2604.14116v22026arXiv
Intern-S1-Pro: Scientific Multimodal Foundation Model at Trillion ScaleYicheng Zou, Dongsheng Zhu, Lin Zhu et al.1 citation2603.25040v22026arXiv
DataChef: Cooking Up Optimal Data Recipes for LLM Adaptation via Reinforcement LearningYicheng Chen, Zerun Ma, Xinchen Xie et al.0 citations2602.11089v22026arXiv