Inference of Affordances and Active Motor Control in Simulated Agents

TL;DR

This paper presents a neural network architecture based on active inference principles that learns affordance maps and enables simulated agents to perform flexible, goal-directed behaviors with strong generalization across diverse environments.

Contribution

The authors introduce a novel neural network model trained to minimize free energy, which develops interpretable affordance maps and supports adaptive, goal-oriented actions in simulation.

Findings

The model learns latent states as affordance maps indicating action effects.

Agents exhibit flexible, goal-directed navigation and obstacle avoidance.

High zero-shot generalization to procedurally generated environments.

Abstract

Flexible, goal-directed behavior is a fundamental aspect of human life. Based on the free energy minimization principle, the theory of active inference formalizes the generation of such behavior from a computational neuroscience perspective. Based on the theory, we introduce an output-probabilistic, temporally predictive, modular artificial neural network architecture, which processes sensorimotor information, infers behavior-relevant aspects of its world, and invokes highly flexible, goal-directed behavior. We show that our architecture, which is trained end-to-end to minimize an approximation of free energy, develops latent states that can be interpreted as affordance maps. That is, the emerging latent states signal which actions lead to which effects dependent on the local context. In combination with active inference, we show that flexible, goal-directed behavior can be invoked, incorporating the emerging affordance maps. As a result, our simulated agent flexibly steers through continuous spaces, avoids collisions with obstacles, and prefers pathways that lead to the goal with high certainty. Additionally, we show that the learned agent is highly suitable for zero-shot generalization across environments: After training the agent in a handful of fixed environments with obstacles and other terrains affecting its behavior, it performs similarly well in procedurally generated environments containing different amounts of obstacles and terrains of various sizes at different locations.