metadata
license: mit
Model card for `boldgpt_small_patch10``
A Vision Transformer (ViT) model trained on BOLD activation maps from NSD-Flat. The training objective was to auto-regressively predict the next patch with shuffled patch order.
license: mit
A Vision Transformer (ViT) model trained on BOLD activation maps from NSD-Flat. The training objective was to auto-regressively predict the next patch with shuffled patch order.