DAGroup-PKU · 7 records

dateidtitleauthorsvenue↑
2026-11-132611.08318Learning the Task or Learning the Verifier? Rethinking Criteria Exposure in Embodied Self-ImprovementWenbo Huang et al.arXiv18
2026-11-122611.16748PhysLoop: Harnessing LLMs' Physics Awareness for Self-Improving Video Generation ModelsYuxuan Chen et al.arXiv46
2026-11-062611.06817Open-HumanScale: An Open Recipe for Scaling Human and Robot Experience for Embodied Foundation ModelsJuncheng Ma et al.arXiv20
2026-11-052611.05320PyRUA-DataFarm: LLM Agent as a Cost-Efficient Teleoperator for Real-Ready Robot DataKewei Zhang et al.arXiv26
2026-10-082610.02368Rethinking World-Action Model for Compositional and In-Context Robotic ManipulationShukai Gong et al.arXiv69
2026-10-022610.01939Fewer Tokens, Better Action: GPT-6 Astra Robot Agents with 14% Higher Success Rate but 65% Fewer TokensRuiyang Si et al.arXiv26
2026-10-012609.39222DC-SAE: Deep Compression Semantic Autoencoder for Faster Diffusion ConvergenceXu Huang et al.arXiv25