SWE-OPD: On-Policy Distillation for Agentic Codingwith Yuhang He, Yanyu Chen, Chuanzhe Guo, Jie Zhu +91 versionW_br36tyzev1 · current
Speculative Rollback Correction for Quality-Diverse Web Agent ImitationLongkun Hao, Hongyu Lin, Hao Li et al.0 citations2606.12485v22026arXiv
Where Hindsight Credit Can Reside: A Signed-Capacity View of Token Updates in RLVRYuhang He, Haodong Wu, Siyi Liu et al.0 citations2604.11056v22026arXiv
GLM-5: from Vibe Coding to Agentic EngineeringGLM-5-Team: Aohan Zeng, Xin Lv, Zhenyu Hou et al.6 citations2602.15763v22026arXiv