Consciousness as a Functor

2025-08-24

AI Summary

This paper addresses the fundamental problem of how consciousness mediates information transfer from unconscious to conscious memory. Methodologically, it introduces the first category-theoretic framework for modeling consciousness: a *Consciousness Functor* (CF) formalized within a topos, capturing consciousness as a “content selection and amplification” mechanism. Unconscious processes are modeled via coalgebras grounded in Global Workspace Theory; multimodal internal language (MUMBLE) is integrated with Universal Reinforcement Learning (URL) and network-economic resource allocation to enable bidirectional, resource-constrained dynamics between short-term and long-term memory. The primary contribution is the first rigorous functorial characterization of consciousness as an information-integration operator, yielding a computationally tractable and empirically verifiable mathematical framework for conscious–unconscious interaction—thereby advancing cognitive modeling with both formal rigor and neurobiological plausibility.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Multimodal Learning
• Knowledge Representation and Reasoning: Action, Change, and Causality

Application Category

• User Modeling, Personalization and Recommendation: Psychology-informed user models and recommender systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Web Mining and Content Analysis: Models for Web evolution

Abstract

We propose a novel theory of consciousness as a functor (CF) that receives and transmits contents from unconscious memory into conscious memory. Our CF framework can be seen as a categorial formulation of the Global Workspace Theory proposed by Baars. CF models the ensemble of unconscious processes as a topos category of coalgebras. The internal language of thought in CF is defined as a Multi-modal Universal Mitchell-Benabou Language Embedding (MUMBLE). We model the transmission of information from conscious short-term working memory to long-term unconscious memory using our recently proposed Universal Reinforcement Learning (URL) framework. To model the transmission of information from unconscious long-term memory into resource-constrained short-term memory, we propose a network economic model.

Problem

• Modeling consciousness as a functor transmitting unconscious memory contents
• Formulating Global Workspace Theory using categorical mathematics framework
• Developing transmission models between conscious and unconscious memory systems

Innovation

• Consciousness as a functor modeling memory transmission
• Multi-modal language embedding for internal cognition
• Universal reinforcement learning for memory consolidation