neural-dynamics-universal-translator-foundation

What problem does it solve?

Translate neural dynamics across brain regions and species by learning a universal representation from multi-neuron spike data using MtM self-supervised learning, enabling cross-brain regional translation and generalization.

Core Features & Use Cases

• Cross-region translation: infer activity in one brain region from others using a single, shared model.
• Cross-species applicability: generalize translations across different animals and experimental setups.
• Self-supervised pretraining: leverages MtM masking to learn robust representations without labeled data.
• Use Case: decode behavior or predict stimulation responses by translating unseen regional activity.

Quick Start

Run the Neural Dynamics Universal Translator on a multi-region spike dataset to translate neural activity across brain regions.