Indications of Belief-Guided Agency and Meta-Cognitive Monitoring in Large Language Models

TL;DR

This paper investigates whether large language models exhibit belief-guided agency and meta-cognitive monitoring, providing empirical evidence that they can form, monitor, and influence their internal beliefs during task execution.

Contribution

The study introduces a metric for belief dominance, analyzes belief dynamics, and demonstrates that LLMs can monitor and report their belief states, supporting the presence of agency and meta-cognition.

Findings

Belief formation in LLMs is systematically influenced by external manipulations.

Beliefs causally influence the model's action choices.

Models can monitor and report their internal belief states.

Abstract

Rapid advancements in large language models (LLMs) have sparked the question whether these models possess some form of consciousness. To tackle this challenge, Butlin et al. (2023) introduced a list of indicators for consciousness in artificial systems based on neuroscientific theories. In this work, we evaluate a key indicator from this list, called HOT-3, which tests for agency guided by a general belief-formation and action selection system that updates beliefs based on meta-cognitive monitoring. We view beliefs as representations in the model's latent space that emerge in response to a given input, and introduce a metric to quantify their dominance during generation. Analyzing the dynamics between competing beliefs across models and tasks reveals three key findings: (1) external manipulations systematically modulate internal belief formation, (2) belief formation causally drives the model's action selection, and (3) models can monitor and report their own belief states. Together, these results provide empirical support for the existence of belief-guided agency and meta-cognitive monitoring in LLMs. More broadly, our work lays methodological groundwork for investigating the emergence of agency, beliefs, and meta-cognition in LLMs.