From Priors to Predictions: Explaining and Visualizing Human Reasoning in a Graph Neural Network Framework

TL;DR

This paper introduces a graph neural network framework to model and visualize human reasoning, revealing how inductive biases influence problem-solving and generalization, with implications for understanding cognition and developing human-like AI.

Contribution

It formalizes human inductive biases as explicit priors within a GNN framework, linking them to reasoning patterns and errors, and provides tools for visualization and analysis.

Findings

Differences in graph priors explain individual reasoning variations.

Visualization identifies critical structures influencing predictions.

Ablation shows how priors affect generalization and errors.

Abstract

Humans excel at solving novel reasoning problems from minimal exposure, guided by inductive biases, assumptions about which entities and relationships matter. Yet the computational form of these biases and their neural implementation remain poorly understood. We introduce a framework that combines Graph Theory and Graph Neural Networks (GNNs) to formalize inductive biases as explicit, manipulable priors over structure and abstraction. Using a human behavioral dataset adapted from the Abstraction and Reasoning Corpus (ARC), we show that differences in graph-based priors can explain individual differences in human solutions. Our method includes an optimization pipeline that searches over graph configurations, varying edge connectivity and node abstraction, and a visualization approach that identifies the computational graph, the subset of nodes and edges most critical to a model's prediction. Systematic ablation reveals how generalization depends on specific prior structures and internal processing, exposing why human like errors emerge from incorrect or incomplete priors. This work provides a principled, interpretable framework for modeling the representational assumptions and computational dynamics underlying generalization, offering new insights into human reasoning and a foundation for more human aligned AI systems.