Zuxuan Wu

AgentGym-RL: An Open-Source Framework to Train LLM Agents for Long-Horizon Decision Making via Multi-Turn RL

VTBench: Comprehensive Benchmark Suite Towards Real-World Virtual Try-on Models

PreferThinker: Reasoning-based Personalized Image Preference Assessment

RoboOmni: Proactive Robot Manipulation in Omni-modal Context

Efficient-LVSM: Faster, Cheaper, and Better Large View Synthesis Model via Decoupled Co-Refinement Attention

CameraNoise: Learning Precise Camera Control with Video Diffusion in Noise Space

UniHand: A Unified Model for Diverse Controlled 4D Hand Motion Modeling

ViaRL: Adaptive Temporal Grounding via Visual Iterated Amplification Reinforcement Learning

TrajTok: What makes for a good trajectory tokenizer in behavior generation?

HiMoE-VLA: Hierarchical Mixture-of-Experts for Generalist Vision–Language–Action Policies

CreatiDesign: A Unified Multi-Conditional Diffusion Transformer for Creative Graphic Design

Daily-Omni: Towards Audio-Visual Reasoning with Temporal Alignment across Modalities

Unleashing 2D Rewards for Human Preference Aligned Text-to-3D Generation via Preference Score Distillation

CaTok: Taming Mean Flows for One-Dimensional Causal Image Tokenization

FlashPlanner: Accelerating Diffusion-based Planner for Autonomous Driving via Globally Consistent Velocity Field and Redundancy Reduction

PointARSim: Point-cloud–enhanced Generative Auto-Regressive Simulation for Closed-loop End-to-end Autonomous Driving Evaluation

ReGen4AD: Retrieval Based Online Video Generation for Reactive Autonomous Driving Simulation

CoSMo-RL: Towards Trustworthy LMRMs via Joint Safety and Stability

ZTRS: Zero-Imitation End-to-end Autonomous Driving with Trajectory Scoring

Compositional Text-to-Image Generation Via Region-aware Bimodal Direct Preference Optimization

FreezeVLA: Action-Freezing Attacks against Vision-Language-Action Models

OmniGen-AR: AutoRegressive Any-to-Image Generation

Adaptive Retention & Correction: Test-Time Training for Continual Learning

Seg2Any: Open-set Segmentation-Mask-to-Image Generation with Precise Shape and Semantic Control

ForgerySleuth: Empowering Multimodal Large Language Models for Image Manipulation Detection

UniGen: Enhanced Training & Test-Time Strategies for Unified Multimodal Understanding and Generation

AgentGym: Evaluating and Evolving Large Language Model-based Agents across Diverse Envronments

INST-IT: Boosting Instance Understanding via Explicit Visual Prompt Instruction Tuning

BlockDance: Reuse Structurally Similar Spatio-Temporal Features to Accelerate Diffusion Transformers

Hydra-MDP++: Advancing End-to-End Driving via Hydra-Distillation with Expert-Guided Decision Analysis

TinyMem: Condensing Multimodal Memory for Long-form Video Action Detection

GeoGS3D: Single-view 3D Reconstruction via Geometric-aware Diffusion Model and Gaussian Splatting