Dinghuai Zhang

FlowRL: Matching Reward Distributions for LLM Reasoning

Generalized Representation for Generalized Dynamics Generation

``Noisier'’ Noise Contrastive Estimation is (Almost) Maximum Likelihood

Coevolutionary Continuous Discrete Diffusion: Make Your Diffusion Language Model a Latent Reasoner

Diffusion Large Language Models for Black-Box Optimization

Self-Evolving Curriculum for LLM Reasoning

Value Gradient Guidance for Flow Matching Alignment

Nabla-R2D3: Effective and Efficient 3D Diffusion Alignment with 2D Rewards

Next Semantic Scale Prediction via Hierarchical Diffusion Language Models

Denoising Autoregressive Transformers for Scalable Text-to-Image Generation

Efficient Diversity-Preserving Diffusion Alignment via Gradient-Informed GFlowNets

EnzymeFlow: Generating Reaction-specific Enzyme Catalytic Pockets through Flow Matching and Co-Evolutionary Dynamics