Xinchao Wang

Invisible Safety Threat: Malicious Finetuning for LLM via Steganography

Flash-DD: An Ultra Parameter-Efficient Approach to Dataset Distillation

dParallel: Learnable Parallel Decoding for dLLMs

Visual Geometry Transformer in the Wild: Distractor-Free 3D Reconstruction

Paying More Attention to Images in Large-scale Dataset Compression

SparseD: Sparse Attention for Diffusion Language Models

PixelThink: Towards Efficient Chain-of-Pixel Reasoning

Efficient Multi-modal Dataset Distillation via Analytic Parameter Matching

Understanding Dataset Distillation via Spectral Filtering

CoDA: From Text-to-Image Diffusion Models to Training-Free Dataset Distillation

Every Step Counts: Decoding Trajectories as Authorship Fingerprints of dLLMs

ConciseHint: Boosting Efficient Reasoning via Continuous Concise Hints during Generation

Dimple: Discrete Diffusion Multimodal Large Language Model with Parallel Decoding

Gradient Inversion Attacks Beyond SGD

Revisit Model Adaptation from Parameters to Features

Privacy-Aware Hybrid Image Synthesis with Local-Cloud Collaboration

Can MLLMs Guide Me Home? A Benchmark Study on Fine-grained Visual Reasoning from Transit Maps

Transferring Jailbreak Attacks from Public to Private LLMs via Local Prompt Optimization

SCoT: Self-Correction at Test-time for Image Generation

Diversity-Guided MLP Pruning for Efficient Large Vision Transformers

Don't Forget the Nonlinearity: Unlocking Activation Functions in Efficient Fine-Tuning

Epipolar Prompt: A Simple Baseline for Motion Segmentation

MixReasoning: Switching Modes to Think

NoLan: Mitigating Object Hallucinations in Large Vision-Language Models via Dynamic Suppression of Language Priors

Sketch First, Scale Fast: On Efficient Inference-Time Scaling for Visual Generation

Disentangling Token Dependencies for Efficient Decoding in Diffusion Language Models

Vid-SME: Membership Inference Attacks against Large Video Understanding Models

Poison-splat: Computation Cost Attack on 3D Gaussian Splatting

Thinkless: LLM Learns When to Think

Ultra-Resolution Adaptation with Ease

Image Editing As Programs with Diffusion Models

GraphBridge: Towards Arbitrary Transfer Learning in GNNs

Asymmetric Dual-Lens Video Deblurring

TreeSplat: Mergeable Tree for Deformable Gaussian Splatting

CLEAR: Conv-Like Linearization Revs Pre-Trained Diffusion Transformers Up

Test3R: Learning to Reconstruct 3D at Test Time

VeriThinker: Learning to Verify Makes Reasoning Model Efficient

Kolmogorov-Arnold Transformer

Compositional Video Generation as Flow Equalization

HG-Adapter: Improving Pre-Trained Heterogeneous Graph Neural Networks with Dual Adapters

Machine Unlearning in 3D Generation: A Perspective-Coherent Acceleration Framework

1000+ FPS 4D Gaussian Splatting for Dynamic Scene Rendering

dKV-Cache: The Cache for Diffusion Language Models

LinFusion: 1 GPU, 1 Minute, 16K Image

Control and Realism: Best of Both Worlds in Layout-to-Image without Training

Neural Phylogeny: Fine-Tuning Relationship Detection among Neural Networks

IFAdapter: Instance feature control for grounded Text-to-Image Generation

Point Cloud Dataset Distillation

Heavy Labels Out! Dataset Distillation with Label Space Lightening

Point Cloud Dataset Distillation

Hash3D: Training-free Acceleration for 3D Generation

Video-Infinity: Distributed Long Video Generation

Adversarial Attacks on Fine-tuned LLMs

Injecting Vision Language into Autoregressive Image Generation