From Kicking to Causality: Simulating Infant Agency Detection with a Robust Intrinsic Reward

TL;DR

This paper introduces the Causal Action Influence Score (CAIS), an intrinsic reward based on causal inference, enabling reinforcement learning agents to robustly identify their influence in noisy environments, akin to infant agency detection.

Contribution

The paper presents CAIS, a novel causal intrinsic reward that improves agent robustness in noisy scenarios, advancing the understanding of causal agency detection in AI systems.

Findings

CAIS enables agents to filter environmental noise and learn correct policies.

Agents with CAIS can reproduce the 'extinction burst' phenomenon.

CAIS outperforms correlation-based rewards in noisy, ecologically valid scenarios.

Abstract

While human infants robustly discover their own causal efficacy, standard reinforcement learning agents remain brittle, as their reliance on correlation-based rewards fails in noisy, ecologically valid scenarios. To address this, we introduce the Causal Action Influence Score (CAIS), a novel intrinsic reward rooted in causal inference. CAIS quantifies an action's influence by measuring the 1-Wasserstein distance between the learned distribution of sensory outcomes conditional on that action, $p(h|a)$, and the baseline outcome distribution, $p(h)$. This divergence provides a robust reward that isolates the agent's causal impact from confounding environmental noise. We test our approach in a simulated infant-mobile environment where correlation-based perceptual rewards fail completely when the mobile is subjected to external forces. In stark contrast, CAIS enables the agent to filter this noise, identify its influence, and learn the correct policy. Furthermore, the high-quality predictive model learned for CAIS allows our agent, when augmented with a surprise signal, to successfully reproduce the "extinction burst" phenomenon. We conclude that explicitly inferring causality is a crucial mechanism for developing a robust sense of agency, offering a psychologically plausible framework for more adaptive autonomous systems.