Probabilistic Foundations for Metacognition via Hybrid-AI

TL;DR

This paper introduces a probabilistic framework for metacognition in AI, specifically through hybrid systems that learn to correct neural models, providing theoretical insights into their capabilities and limitations.

Contribution

It develops a rigorous probabilistic foundation for metacognitive error correction in hybrid-AI systems, extending prior empirical work with formal proofs and conditions.

Findings

Proves necessary and sufficient conditions for metacognitive improvement

Identifies limits of error detecting and correcting rules in hybrid-AI

Provides a formal probabilistic framework for future research

Abstract

Metacognition is the concept of reasoning about an agent's own internal processes, and it has recently received renewed attention with respect to artificial intelligence (AI) and, more specifically, machine learning systems. This paper reviews a hybrid-AI approach known as "error detecting and correcting rules" (EDCR) that allows for the learning of rules to correct perceptual (e.g., neural) models. Additionally, we introduce a probabilistic framework that adds rigor to prior empirical studies, and we use this framework to prove results on necessary and sufficient conditions for metacognitive improvement, as well as limits to the approach. A set of future