Guy Van den Broeck

Planned Diffusion

Learning Tractable Distributions of Language Model Continuations

The Pitfalls of KV Cache Compression

Learning to Discretize Denoising Diffusion ODEs

Accelerating Diffusion LLMs via Adaptive Parallel Decoding

Rao-Blackwell Gradient Estimators for Equivariant Denoising Diffusion

Plug-and-Play Context Feature Reuse for Efficient Masked Generation

The Limits of Tractable Marginalization

Tractable Transformers for Flexible Conditional Generation

TRACE Back from the Future: A Probabilistic Reasoning Approach to Controllable Language Generation

Controllable Generation via Locally Constrained Resampling

Discrete Copula Diffusion

Scaling Probabilistic Circuits via Monarch Matrices