Transform brain data into AI-ready insights.
EEG-DaSh (Data Share) is an open-access platform democratizing MEEG (EEG, MEG, iEEG, and MEG) data for machine learning, deep learning, and computational neuroscience.
Why another data portal and tool? Because 25% of public EEG datasets are unusable and most require weeks, months to prepare for ML.
We do the hard work-preprocessing, standardization, feature engineering- so you can focus on science.
- ML-Ready Data: Preprocessed EEG/MEG in PyTorch-compatible formats data wrangling required
- Smart Integration: Direct access to NSF supercomputers via the [Neuroscience Gateway]
-
- Community-Driven: Built on BIDS standards with data from OpenNeuro, NEMAR, and contributors worldwide
-
- Hugging Face Compatibility: Compatibility with Hugging Face for easy transfer of pre-processed data.
Currently hosting diverse MEEG datasets preprocessed for immediate use. All data:
- Stored in standardized, documented formats
- Linked back to original sources (OpenNeuro, peer-reviewed papers)
- Annotated with rich experimental metadata
- Ready for aggregation across studies
Led by: University of California, San Diego (UCSD) & Ben-Gurion University of the Negev (BGU)
Funded by: National Science Foundation (NSF)
Status: Beta-we're building this in the open. Join us!
From Cross-Task to Cross-Subject EEG Decoding
A competition pushing the frontiers of EEG transfer learning:
- 3,000+ participants, 6 cognitive tasks
- $7,500 in prizes + travel grants + conference registration
- Challenge details & starter kits | ArXiv paper
- Data Formats: BIDS (standardization) → BrainDecode (ML-ready)
- Processing: BrainDecode (PyTorch), EEGPrep, MNE-Python
- Browse datasets: Visit the portal
- Code examples: PyTorch Lightning + Google Colab notebooks included
- Database Paper: EEG-DaSh: An Open Data, Tool, and Compute Resource (in review)
- Docs & Tutorials: Full documentation
 UC San Diego |
 Ben-Gurion Univ. |
Status: Beta ⚡ | License: BSD 3-Clause | Welcome: All contributions & feedback!
Got data to share? See our Github page. Have feedback? Open an issue.