Chinese University of Hong Kong, Shenzhen · 11 records

dateidtitleauthorsvenue↑
2026-11-062611.06982Decode Before You Hear: Speculative Listening for Chunk-Interleaved Full-Duplex Speech Language ModelsYutong Bian et al.arXiv5
2026-10-222610.22740R^2Orch: Recursive Self-Improvement for Agent Orchestration via Self-Play within a Skill-Internalization LoopMingda Zhang et al.arXiv52
2026-10-142610.12412WorldCast: Distributed Multiplayer World ModelsZiyang Ye et al.arXiv4
2026-10-132610.12099UNITAS: A 3D-Native World Action Model for Embodied ManipulationRuixiang Wang et al.arXiv7
2026-10-072610.05966HuatuoGPT-3: RL-Only Domain Adaptation from Base ModelsJunying Chen et al.arXiv41
2026-09-142609.11682COBRA-Skills: Contextual Bandit-Guided Evolution for Agent Skill OptimizationPingchen Lu et al.arXiv43
2026-09-032609.02886SolarWM: Open Data and Scalable Training for Long-Horizon Video World ModelsJunchao Huang et al.arXiv115
2026-08-112608.04419SPOT: Sparse Probing and Outcome Calibration for On-Policy DistillationZikun Qu et al.arXiv15
2026-07-062607.02471Interpretation-Oriented Cloud Removal via Observation-Anchored Residual Flow with Geo-Contextual AlignmentZiyao Wang et al.arXiv2
2026-06-172606.17861GameCraft-Bench: Can Agents Build Playable Games End-to-End in a Real Game Engine?Tongxu Luo et al.arXiv58
2026-06-112606.09323TRL-Bench: Standardizing Cross-Paradigm Representation-Level Evaluation of Tabular EncodersWei Pang et al.arXiv54