Yutong Bai

The Serial Scaling Hypothesis

GRAID: Enhancing Spatial Reasoning of VLMs through High-Fidelity Data Generation

Transformers Adaptively Learn Molecular Structures Without Graph Priors

AV-Odyssey Bench: From Fundamental Audio Perception to Audio-Visual Understanding

Adaptively Labeling Vision Datasets Via Instance-Level Retrieval

KiVA: Kid-inspired Visual Analogies for Testing Large Multimodal Models

Whole-Body Conditioned Egocentric Video Prediction

REOrdering Patches Improves Vision Models

Analyzing the Language of Visual Tokens

ACAV-1M: Data Curation and Benchmarking for Audio-Visual Representation Learning