Error Correction

 Okay, so Algonauts plan has matured. I will implement an encoder based on a pre-trained Contrastive Learning model and Partial Least Squares Regression. And then I will implement my own innovation! A PCA error correction step:-) So I will learn the PC's of voxels on the training data and project my predictions on these PC's as error correction:-)

And I'll make visualizations and diagnostics in StreamLit! And write about it in my PhD Thesis:-)