Chongxuan Li

Variational Reasoning for Language Models

UltraLLaDA: Scaling the Context Length to 128K for Diffusion Large Language Models

Principled RL for Diffusion LLMs Emerges from a Sequence-Level Perspective

UltraViCo: Breaking Extrapolation Limits in Video Diffusion Transformers

ReFusion: A Diffusion Large Language Model with Parallel Autoregressive Decoding

Reinforcing General Reasoning Without Verifiers

LLaDA 1.5: Variance-Reduced Preference Optimization for Large Language Diffusion Models

Large Language Diffusion Models

Masked Diffusion Models as Energy Minimization

RIFLEx: A Free Lunch for Length Extrapolation in Video Diffusion Transformers

A Theory for Conditional Generative Modeling on Multiple Data Sources

Scaling Diffusion Transformers Efficiently via $\mu$P

FlexWorld: Progressively Expanding 3D Scenes for Flexible-View Exploration

Scaling up Masked Diffusion Models on Text

Your Absorbing Discrete Diffusion Secretly Models the Conditional Distributions of Clean Data

Effective and Efficient Masked Image Generation Models

MissDiff: Training Diffusion Models on Tabular Data with Missing Values

Improving Long-Text Alignment for Text-to-Image Diffusion Models

Are Images Indistinguishable to Humans Also Indistinguishable to Classifiers?

Large Language Diffusion Models

MAC: A Multimodal Benchmark for Understanding and Generating Academic Journal Covers