Abstract: High-fidelity synthetic tabular data generation is crucial, given that downstream tasks and data analysis typically require large amounts of tabular data. However, existing methods struggle ...
Modality-agnostic decoders leverage modality-invariant representations in human subjects' brain activity to predict stimuli irrespective of their modality (image, text, mental imagery).