|
Download README.md from Dcode-AI/SAMBA-2s: direct link, hf CLI and curl.
- Browser
- Download file 3.7 kB
-
https://huggingface.co/Dcode-AI/SAMBA-2s/resolve/main/README.md
- Command line
-
hf download hf://Dcode-AI/SAMBA-2s/README.md
-
curl -L -o README.md https://huggingface.co/Dcode-AI/SAMBA-2s/resolve/main/README.md
3.7 kB
| pipeline_tag: feature-extraction | |
| tags: | |
| - eeg | |
| - brain-computer-interface | |
| - self-supervised-learning | |
| - mamba | |
| - pytorch | |
| # SAMBA-2s | |
| [Project website](https://samba.dcode.ai/) 路 [Code](https://github.com/Jiazhen-Hong/SAMBA) 路 [Paper](https://arxiv.org/abs/2511.18571) 路 [SAMBA-100s](https://huggingface.co/Dcode-AI/SAMBA-100s) | |
| SAMBA-2s was pretrained on 2-second EEG windows from Emotiv (Alpha, Attention, Crowdsource, DriverDistraction, and STEW), HBN resting state, Muse sleep onset, and DICE eyes closed. Pretraining used 256 samples per window at 128 Hz and 4, 14, 19, and 129 EEG channels. The encoder supports other window lengths and electrode layouts. | |
| ## Download | |
| [Download config](https://huggingface.co/Dcode-AI/SAMBA-2s/resolve/main/config.json?download=true) 路 [Download weights](https://huggingface.co/Dcode-AI/SAMBA-2s/resolve/main/model.safetensors?download=true) | |
| Keep both files in the same directory. To download them together from a terminal: | |
| ```bash | |
| pip install -U huggingface_hub | |
| hf download Dcode-AI/SAMBA-2s config.json model.safetensors --local-dir ./SAMBA-2s | |
| ``` | |
| ## Usage | |
| Install SAMBA in a CUDA environment: | |
| ```bash | |
| git clone https://github.com/Jiazhen-Hong/SAMBA.git | |
| cd SAMBA | |
| pip install -e '.[mamba,hub]' | |
| ``` | |
| Load the downloaded directory. If you skipped the download step, the following example fetches both files automatically: | |
| ```python | |
| import torch | |
| from huggingface_hub import snapshot_download | |
| from samba_eeg import SambaModel | |
| folder = snapshot_download( | |
| "Dcode-AI/SAMBA-2s", | |
| allow_patterns=["config.json", "model.safetensors"], | |
| ) | |
| model = SambaModel.from_pretrained(folder).cuda().eval() | |
| ``` | |
| For files you already downloaded, pass their local directory to `SambaModel.from_pretrained` in place of `folder`. | |
| Prepare EEG at 128 Hz. Use `signals` shaped `[batch, channels, samples]` and electrode `positions` shaped `[channels, 3]` or `[batch, channels, 3]`. Match the signal channel order and use the same coordinate frame and units as the target positions in `config.json`. The decoder reconstructs EEG at the input electrode coordinates. | |
| ```python | |
| signals = signals.to(device="cuda", dtype=torch.float32) | |
| positions = positions.to(device="cuda", dtype=torch.float32) | |
| # Skip these two lines if the signals are already normalized this way. | |
| scale = torch.quantile(signals.abs(), 0.95, dim=-1, keepdim=True) | |
| signals = signals / scale.clamp_min(1e-8) | |
| with torch.inference_mode(): | |
| features = model.encode(signals, positions) | |
| pooled_features = features.mean(dim=1) | |
| ``` | |
| `features` has shape `[batch, floor(samples / 8), 128]`. Use `pooled_features` for linear probing. | |
| ## Learned EEG spatial weights | |
| SAMBA uses electrode coordinates to learn how to combine EEG channels. These maps show how the spatial weights change during training. | |
| | Eyes open vs. closed | Driver distraction | Motor imagery | | |
| |:---:|:---:|:---:| | |
| | <img src="https://raw.githubusercontent.com/Jiazhen-Hong/SAMBA/main/Figures/Topo_gif/EyesState_GIF.gif" width="250" alt="Eyes open vs. closed"> | <img src="https://raw.githubusercontent.com/Jiazhen-Hong/SAMBA/main/Figures/Topo_gif/DriverDistraction_GIF.gif" width="238" alt="Driver distraction"> | <img src="https://raw.githubusercontent.com/Jiazhen-Hong/SAMBA/main/Figures/Topo_gif/MI_GIF.gif" width="245" alt="Motor imagery"> | | |
| ## Citation | |
| ```bibtex | |
| @inproceedings{hong2026samba, | |
| title = {{SAMBA}: Toward an {EEG} Foundation Model via Differential Mamba and Coordinate-Driven Spatial Embedding}, | |
| author = {Hong, Jiazhen and Mackellar, Geoffrey and Ghane, Soheila}, | |
| booktitle = {NeurReps 2026: Symmetry and Geometry in Neural Representations (Proceedings Track)}, | |
| year = {2026} | |
| } | |
| ``` | |