The Dataset Viewer has been disabled on this dataset.

TabBioMed benchmark results

Evaluation results for the TabBioMed living benchmark, including metrics, measured timings, test predictions and hyperparameter-search histories. The benchmark datasets are available in TabBioMed/TabBioMed.

Layout

Each <Model>/<Dataset>/ folder contains:

  • report.json and timing.json: final evaluation outcomes and measured costs.
  • Prediction .pt files and test-label .npy files for available evaluations.
  • hpo/: recorded configurations, validation metrics, and selections at 25/50/100/200 attempted configurations, with additional saved test predictions.
  • pca_500/ and a100_retry/: the corresponding separate evaluations, where applicable.

Cross-validation records identify their outer folds. Missing evaluations remain explicit in the reports. Prediction references are relative to their containing hpo/ folder and may point to an existing prediction in the parent folder.

Download

from huggingface_hub import snapshot_download

results = snapshot_download(
    repo_id="TabBioMed/results",
    repo_type="dataset",
    local_dir="results",
)

For a fixed release, pass its commit hash as revision. Select particular models or file types with allow_patterns. While the repository is private, authenticate with a Hugging Face account that has access.

Citation

If you use TabBioMed in your work, please cite:

@inproceedings{tabbiomed2026,
  title     = {TabBioMed: A Large-Scale Benchmark for Biomedical Tabular Learning},
  author    = {Mateo-Bernad{\'o}, Pau and Polapragada, Saivenkata Nagavyjayanthi and Bonet, David and Arbiol Rakuljic, Pol and Mogas Pladevall, Laia and Comajoan Cara, Mar{\c{c}}al and Gonzalez-Ferrer, Jesus and Abante, Jordi and Mas Montserrat, Daniel and Haussler, David and Ioannidis, Alexander G.},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS), Evaluations and Datasets Track},
  year      = {2026}
}
Downloads last month
27