Neuropsychology of AI: Relationship Between Activation Proximity and Categorical Proximity Within Neural Categories of Synthetic Cognition

TL;DR

This paper explores how concepts from neuropsychology, especially categorization, can enhance the interpretability of artificial neural networks by examining the relationship between activation proximity and categorical proximity in synthetic cognition.

Contribution

It introduces a neuropsychological framework to analyze neural categories in AI, linking activation patterns to cognitive categorization processes for better explainability.

Findings

Activation proximity correlates with categorical proximity in neural representations.

Neural categories in AI can be interpreted through cognitive psychology concepts.

The approach improves understanding of neural network segmentation and construction of reality.

Abstract

Neuropsychology of artificial intelligence focuses on synthetic neural cog nition as a new type of study object within cognitive psychology. With the goal of making artificial neural networks of language models more explainable, this approach involves transposing concepts from cognitive psychology to the interpretive construction of artificial neural cognition. The human cognitive concept involved here is categorization, serving as a heuristic for thinking about the process of segmentation and construction of reality carried out by the neural vectors of synthetic cognition.