Michael Eickenberg

PETRA: Parallel End-to-end Training with Reversible Architectures

From Feature Visualization to Visual Circuits: Effect of Model Perturbation

Multiple Physics Pretraining for Spatiotemporal Surrogate Models

Multiple Physics Pretraining for Physical Surrogate Models

xVal: A Continuous Number Encoding for Large Language Models

PETRA: Parallel End-to-end Training with Reversible Architectures

MoMo: Momentum Models for Adaptive Learning Rates