Know Thyself by Knowing Others: Learning Neuron Identity from Population Context

2025-11-30

AI Summary

This work addresses the neuron identity inference problem—i.e., predicting cell type, brain region assignment, and connectivity properties from population neural activity—by proposing NuCLR, a self-supervised framework. Methodologically, it introduces a permutation-equivariant spatiotemporal Transformer architecture that integrates multimodal neural recordings (calcium imaging and electrophysiology) via contrastive learning, modeling population context in a permutation-invariant manner, and systematically investigates scaling laws in neuron representation learning for the first time. Key contributions include: (1) enabling cross-animal zero-shot transfer, substantially improving label efficiency; and (2) achieving state-of-the-art performance in cell-type and brain-region decoding across multiple benchmark datasets, with strong generalization attained using only minimal labeled data.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Multimodal Learning
• Computer Vision: Representation Learning for Vision

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Neurons process information in ways that depend on their cell type, connectivity, and the brain region in which they are embedded. However, inferring these factors from neural activity remains a significant challenge. To build general-purpose representations that allow for resolving information about a neuron's identity, we introduce NuCLR, a self-supervised framework that aims to learn representations of neural activity that allow for differentiating one neuron from the rest. NuCLR brings together views of the same neuron observed at different times and across different stimuli and uses a contrastive objective to pull these representations together. To capture population context without assuming any fixed neuron ordering, we build a spatiotemporal transformer that integrates activity in a permutation-equivariant manner. Across multiple electrophysiology and calcium imaging datasets, a linear decoding evaluation on top of NuCLR representations achieves a new state-of-the-art for both cell type and brain region decoding tasks, and demonstrates strong zero-shot generalization to unseen animals. We present the first systematic scaling analysis for neuron-level representation learning, showing that increasing the number of animals used during pretraining consistently improves downstream performance. The learned representations are also label-efficient, requiring only a small fraction of labeled samples to achieve competitive performance. These results highlight how large, diverse neural datasets enable models to recover information about neuron identity that generalize across animals. Code is available at https://github.com/nerdslab/nuclr.

Problem

• Inferring neuron identity from neural activity is challenging
• Learning representations to differentiate neurons across stimuli and times
• Achieving generalization in cell type and brain region decoding

Innovation

• Self-supervised contrastive learning for neuron identity
• Spatiotemporal transformer with permutation-equivariant integration
• Scalable pretraining across diverse datasets for generalization