KAST-BAR builds an EEG foundation model by dynamically modeling non-Euclidean brain topology with DSHA and aligning signals to expert semantic space via KASP and STAR, achieving superior performance on six downstream tasks after pre-training on 21 datasets.
The model is trained using the AdamW optimizer ( lr= 5e −5) with a batch size of 128 for 50 epochs
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
KAST-BAR: Knowledge-Anchored Semantically-Dynamic Topology Brain Autoregressive Modeling for Universal Neural Interpretation
KAST-BAR builds an EEG foundation model by dynamically modeling non-Euclidean brain topology with DSHA and aligning signals to expert semantic space via KASP and STAR, achieving superior performance on six downstream tasks after pre-training on 21 datasets.