Brains and algorithms partially converge in natural language processing.

However, the resulting brain scores are i) highly significant (all p < 10 −9 on average across both all fMRI voxels and MEG sensors), including when compared to a permutation baseline (Supplementary Fig. 3 ), and ii) in the same order of magnitude than a baseline shared-response model (or noise ceiling, Fig. 2 ) as well as previous reports (see e.g., 44 before correcting for the noise ceiling).

also in 132,142 other papers