# Unsupervised predictive coding models may explain visual brain representation

## TL;DR

This study investigates whether unsupervised predictive coding models, specifically PredNet, can better predict visual brain activity than supervised models, showing promising results with fMRI and MEG data.

## Contribution

The paper demonstrates that unsupervised predictive coding models trained on video prediction outperform supervised image classifiers in predicting brain activity.

## Key findings

- Unsupervised models achieved 16.67% on fMRI data.
- Unsupervised models achieved 27.67% on MEG data.
- Predictive coding models may better explain visual brain representations.

## Abstract

Deep predictive coding networks are neuroscience-inspired unsupervised learning models that learn to predict future sensory states. We build upon the PredNet implementation by Lotter, Kreiman, and Cox (2016) to investigate if predictive coding representations are useful to predict brain activity in the visual cortex. We use representational similarity analysis (RSA) to compare PredNet representations to functional magnetic resonance imaging (fMRI) and magnetoencephalography (MEG) data from the Algonauts Project. In contrast to previous findings in the literature (Khaligh-Razavi &Kriegeskorte, 2014), we report empirical data suggesting that unsupervised models trained to predict frames of videos may outperform supervised image classification baselines. Our best submission achieves an average noise normalized score of 16.67% and 27.67% on the fMRI and MEG tracks of the Algonauts Challenge.