01

See What You Are Told: Visual Attention Sink in Large Multimodal Models

02

WAVE: Warp-Based View Guidance for Consistent Novel View Synthesis Using a Single Image

03

Spatial Transport Optimization by Repositioning Attention Map for Training-Free Text-to-Image Synthesis

04

Your Large Vision-Language Model Only Needs A Few Attention Heads for Visual Grounding

05

Distilling Spectral Graph for Object-Context Aware Open-Vocabulary Semantic Segmentation

06

Interpreting vision transformers via residual replacement model

07

Rare Text Semantics Were Always There in Your Diffusion Transformer

08

Information-Theoretic Discrete Diffusion

9

SAFEPATH: Preventing Harmful Reasoning in Chain-of-Thought via Early Alignment

10

Assigning Distinct Roles to Quantized and Low-Rank Matrices Toward Optimal Weight Decomposition