Meta FAIR Releases NeuralSet: A Python Package for Neuro-AI That Supports fMRI, M/EEG, Spikes, and HuggingFace Embeddings

Meta FAIR introduced NeuralSet, a Python framework for neuro-AI workflows.

It stores experiments as lightweight event metadata, then uses modality-specific extractors to load and align brain signals with stimulus embeddings.

The design supports fMRI, M/EEG, and spike recordings, with integration paths around tools like Nilearn, MNE-Python, HuggingFace, OpenNeuro, and BIDS.

By decoupling metadata from raw data, NeuralSet aims to reduce friction in building scalable multimodal neuroscience pipelines.