Interpretability in Linear Brain Decoding