LGAIOct 21, 2022

Learning Robust Dynamics through Variational Sparse Gating

DeepMindMILAU of Toronto
arXiv:2210.11698v111 citationsh-index: 26
Originality Highly original
AI Analysis

This addresses the challenge of applying world models to complex, multi-object environments for improved sample-efficiency in AI agents.

The paper tackles the problem of learning world models in environments with many objects by introducing a sparse interactions inductive bias into latent dynamics, resulting in clear improvements over prior models in the BringBackShapes environment.

Learning world models from their sensory inputs enables agents to plan for actions by imagining their future outcomes. World models have previously been shown to improve sample-efficiency in simulated environments with few objects, but have not yet been applied successfully to environments with many objects. In environments with many objects, often only a small number of them are moving or interacting at the same time. In this paper, we investigate integrating this inductive bias of sparse interactions into the latent dynamics of world models trained from pixels. First, we introduce Variational Sparse Gating (VSG), a latent dynamics model that updates its feature dimensions sparsely through stochastic binary gates. Moreover, we propose a simplified architecture Simple Variational Sparse Gating (SVSG) that removes the deterministic pathway of previous models, resulting in a fully stochastic transition function that leverages the VSG mechanism. We evaluate the two model architectures in the BringBackShapes (BBS) environment that features a large number of moving objects and partial observability, demonstrating clear improvements over prior models.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes