Andrew Gordon Wilson

Understanding Transformers for Time Series: Rank Structure, Flow-of-ranks, and Compressibility

A Unification of Discrete, Gaussian, and Simplicial Diffusion

Understanding the Implicit Biases of Design Choices for Time Series Foundation Models

Shrinking Proteins with Diffusion

Zero-shot Forecasting by Simulation Alone

Diverse Dictionary Learning

Understanding and Relaxing the Limitations of Transformers for Linear Algebra

Robust Transfer for Bayesian Optimization with Multi-Task Prior-Fitted Networks

Time-Aware Prior Fitted Networks for Zero-Shot Forecasting with Exogenous Variables

When Does Multimodality Lead to Better Time Series Forecasting?

In-Context Clustering with Large Language Models

Scaling Collapse Reveals Universal Dynamics in Compute-Optimally Trained Neural Networks

Small Batch Size Training for Language Models: When Vanilla SGD Works, and Why Gradient Accumulation is Wasteful

Bayesian Optimization of Antibodies Informed by a Generative Model of Evolving Sequences

How to Scale Second-Order Optimization

Customizing the Inductive Biases of Softmax Attention using Structured Matrices

Improving Discrete Diffusion with Schedule-Conditioning

Enhancing Foundation Models for Time Series Forecasting via Wavelet-based Tokenization

Compute-Optimal LLMs Provably Generalize Better with Scale

Why Masking Diffusion Works: Condition on the Jump Schedule for Improved Discrete Diffusion

Training Flexible Models of Genetic Variant Effects from Functional Annotations using Accelerated Linear Algebra

Enhancing Foundation Models for Time Series Forecasting via Wavelet-based Tokenization

The Promises and Pitfalls of Language Models for Structured Numerical Data

Just How Flexible are Neural Networks in Practice?

Pathologies of Out-of-Distribution Detection

Molecular Active Learning: How can LLMs Help?

LLMs Boost the Performance of Decision Trees on Tabular Data across Sample Sizes