Environmental drivers of systematicity and generalization in a situated agent

TL;DR

This paper investigates how environmental factors influence the ability of neural network-based agents to generalize systematically in a 3D environment, highlighting the importance of diverse experiences and perceptual invariances.

Contribution

It introduces a generic agent architecture and identifies key environmental and training factors that enhance out-of-sample generalization in neural networks.

Findings

More training experiences improve generalization.

Visual invariances from perspective aid generalization.

Variety in perceptual input enhances systematic learning.

Abstract

The question of whether deep neural networks are good at generalising beyond their immediate training experience is of critical importance for learning-based approaches to AI. Here, we consider tests of out-of-sample generalisation that require an agent to respond to never-seen-before instructions by manipulating and positioning objects in a 3D Unity simulated room. We first describe a comparatively generic agent architecture that exhibits strong performance on these tests. We then identify three aspects of the training regime and environment that make a significant difference to its performance: (a) the number of object/word experiences in the training set; (b) the visual invariances afforded by the agent's perspective, or frame of reference; and (c) the variety of visual input inherent in the perceptual aspect of the agent's perception. Our findings indicate that the degree of generalisation that networks exhibit can depend critically on particulars of the environment in which a given task is instantiated. They further suggest that the propensity for neural networks to generalise in systematic ways may increase if, like human children, those networks have access to many frames of richly varying, multi-modal observations as they learn.