A Hardware-oriented Approach for Efficient Active Inference Computation and Deployment

TL;DR

This paper introduces a hardware-efficient methodology for deploying Active Inference algorithms, significantly reducing latency and memory usage to enable real-time and embedded applications.

Contribution

It presents a unified, sparse computational graph that enhances the efficiency of Active Inference deployment on resource-constrained hardware.

Findings

Latency reduced by over 2x

Memory usage decreased by up to 35%

Enables real-time and embedded applications of AIF

Abstract

Active Inference (AIF) offers a robust framework for decision-making, yet its computational and memory demands pose challenges for deployment, especially in resource-constrained environments. This work presents a methodology that facilitates AIF's deployment by integrating pymdp's flexibility and efficiency with a unified, sparse, computational graph tailored for hardware-efficient execution. Our approach reduces latency by over 2x and memory by up to 35%, advancing the deployment of efficient AIF agents for real-time and embedded applications.