Benchmarks

Frozen linear probes, 5-fold subject-level cross-validation. Balanced accuracy (%). These tables are from the previous generation: EPI-250k (base model) and its retired Matryoshka distillation at each truncation dimension. The model behind the API today, EPI-300K (250,000 h real plus 50,000 h synthetic seizure data), scores a 3-seed mean of 72.5% balanced accuracy across the nine internal clinical tasks; its full per-task table ships with the paper.

​

Private clinical tasks

40,909 annotated 30-second epochs from the Swiss Epilepsy Center.

​

Public benchmarks

10 standard public EEG datasets, evaluated under identical conditions. The numeric column headers (768, 384, …) are the truncation dimensions of the retired distilled model.

​

Dimension retention

Mean delta vs the EPI-250k base model, across all 20 tasks. Binary tasks retain accuracy best. Fine-grained multi-class tasks (TUEV, sublabels) and tasks with large domain shift from pre-training data (Parkinsons, MI) degrade more sharply.