The Chinese University of Hong Kong · 14 records

dateidtitleauthorsvenue↑
2026-11-182611.20992MedJev: A System One Vision-Language Decision Model for Fast and Effective Medical Image UnderstandingFan Zhang et al.arXiv10
2026-11-102611.14309MMDM: Measuring Massive Multitask Decision MakingJunhao Shen et al.arXiv27
2026-10-262610.27782RoboJudge: Multimodal Language Models as Judges for Embodied Video GenerationSiyuan Ma et al.arXiv9
2026-10-142610.15772RSIPO: Recursive Self-Improving Policy Optimization for Agentic Reinforcement Learning via Skill-Policy Co-EvolutionFan Zhang et al.arXiv7
2026-09-302609.36636What Makes Recurrence Effective in Looped Language Models?Xinlin Zhuang et al.arXiv7
2026-09-282609.30216Jev in the Wild: A Data-Driven Analysis of the Jev Model's Functionality, Applications and EcosystemGuoming Ling et al.arXiv15
2026-09-242609.23784PackLab: A Comprehensive Framework for Developing, Training, and Evaluating MLLMs in Robotic Bin PackingDonghao Zhou et al.arXiv13
2026-08-182608.14718VideoGAIA: A Benchmark for General AI Assistants on Agentic Video UnderstandingFan Zhang et al.arXiv11
2026-07-152607.11594MAGIC: Transition-Aware Generation of Navigable Multi-Scene Game Worlds with Large Language ModelsTsz Hei Fan et al.arXiv3
2026-06-152606.13707Orchestra-o1: Omnimodal Agent OrchestrationFan Zhang et al.arXiv48
2026-06-102606.07689Struct-Searcher: Agentic Structural Thinking Advances Multimodal Deep Information SeekingFan Zhang et al.arXiv7
2026-05-272605.25802Rethinking VLM Representation for VLA InitializationWeifeng Lin et al.arXiv9
2026-05-192605.17242From Runnable to Shippable: Multi-Agent Test-Driven Development for Generating Full-Stack Web Applications from RequirementsYuxuan Wan et al.arXiv10
2026-05-132605.12492Pion: A Spectrum-Preserving Optimizer via Orthogonal Equivalence TransformationKexuan Shi et al.arXiv4