Jun Zhang

QVGen: Pushing the Limit of Quantized Video Generative Models

WebWeaver: Structuring Web-Scale Evidence with Dynamic Outlines for Open-Ended Deep Research

Low-Latency Neural LiDAR Compression with 2D Context Models

PixelCraft: A Multi-Agent system for High-Fidelity Visual Reasoning on Structured Images

Token-level Data Selection for Safe LLM Fine-tuning

SenseFlow: Scaling Distribution Matching for Flow-based Text-to-Image Distillation

Forge4D: Feed-Forward 4D Human Reconstruction and Interpolation from Uncalibrated Sparse Videos

LumiTex: Towards High-Fidelity PBR Texture Generation with Illumination Context

Towards Efficient Post-Training Quantization For Large Vision-Language Models Via Token-Wise Redundancy Elimination

GAS: Enhancing Reward-Cost Balance of Generative Model-assisted Offline Safe RL

Getting Your LLMs Ready for Reinforcement Learning with Lightweight SFT

UniFork: Exploring Modality Alignment for Unified Multimodal Understanding and Generation

Feed-Forward 3D Gaussian Splatting Compression with Long-Context Modeling

MambaSIC: Mamba-based Stereo Image Compression with Bi-directional Multi-reference Entropy Model

RemedyGS: Defend 3D Gaussian Splatting Against Computation Cost Attack

GI-GS: Global Illumination Decomposition on Gaussian Splatting for Inverse Rendering

Exponential Topology-enabled Scalable Communication in Multi-agent Reinforcement Learning

HarmoniCa: Harmonizing Training and Inference for Better Feature Cache in Diffusion Transformer Acceleration

CAS-Spec: Cascade Adaptive Self-Speculative Decoding for On-the-Fly Lossless Inference Acceleration of LLMs

ReconX: Reconstruct Any Scene from Sparse Views with Video Diffusion Model

C2IQL: Constraint-Conditioned Implicit Q-learning for Safe Offline Reinforcement Learning

AdaWorld: Learning Adaptable World Models with Latent Actions

EVA-Gaussian: 3D Gaussian-based Real-time Human Novel View Synthesis under Diverse Camera Settings

Incremental Aggregated Asynchronous SGD for Arbitrarily Heterogeneous Data

HarmoniCa: Harmonizing Training and Inference for Better Feature Caching in Diffusion Transformer Acceleration

Exploring Selective Layer Fine-Tuning in Federated Learning

MEGA: Memory-Efficient 4D Gaussian Splatting for Dynamic Scenes