In Search of Grandmother Cells: Tracing Interpretable Neurons in Tabular Representations

2026-01-07

arXiv.org

AI Summary

This study investigates whether foundation models contain interpretable neurons analogous to “grandmother cells”—neurons exhibiting high selectivity and salience for a single high-level concept. To this end, we propose two information-theoretic metrics to systematically evaluate neuronal selectivity and salience with respect to specific concepts, and conduct a large-scale search within the representation space of TabPFN. Our work reports the first discovery of statistically significant interpretable neurons in a representation learning model, demonstrating that such neurons can emerge naturally and be identified without complex post-processing. These findings offer a novel perspective and an effective pathway toward enhancing model interpretability.

Technology Category

• Natural Language Processing: Interpretability, Analysis, and Evaluation of NLP Models
• Machine Learning: Deep Neural Architectures and Foundation Models
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Foundation models are powerful yet often opaque in their decision-making. A topic of continued interest in both neuroscience and artificial intelligence is whether some neurons behave like grandmother cells, i.e., neurons that are inherently interpretable because they exclusively respond to single concepts. In this work, we propose two information-theoretic measures that quantify the neuronal saliency and selectivity for single concepts. We apply these metrics to the representations of TabPFN, a tabular foundation model, and perform a simple search across neuron-concept pairs to find the most salient and selective pair. Our analysis provides the first evidence that some neurons in such models show moderate, statistically significant saliency and selectivity for high-level concepts. These findings suggest that interpretable neurons can emerge naturally and that they can, in some cases, be identified without resorting to more complex interpretability techniques.

Problem

• grandmother cells
• interpretable neurons
• tabular foundation models
• neuronal selectivity
• concept representation

Innovation

• grandmother cells
• interpretable neurons
• information-theoretic measures
• TabPFN
• neuronal selectivity