3 papers
cs.LG2025
Soft Geometric Inductive Bias for Object Centric Dynamics
Hampus Linander, Conor Heins, Alexander Tschantz +2
Equivariance is a powerful prior for learning physical dynamics, yet exact group equivariance can degrade performance if the symmetries are broken. We propose object-centric world…
cs.AI2025
AXIOM: Learning to Play Games in Minutes with Expanding Object-Centric Models
Conor Heins, Toon Van de Maele, Alexander Tschantz +11
Current deep reinforcement learning (DRL) approaches achieve state-of-the-art performance in various domains, but struggle with data efficiency compared to human learning, which le…
stat.ML2025
AutoBayes: A Compositional Framework for Generalized Variational Inference
Toby St Clere Smithe, Marco Perin
We introduce a new compositional framework for generalized variational inference, clarifying the different parts of a model, how they interact, and how they compose. We explain tha…