Overview of the datasets, computational models, and analysis. (IMAGE)
Caption
To investigate model-brain alignment in discourse comprehension, we pretrained two BERT-based models and used two neuroimaging datasets. To extract model-based sentence representations, we fed the two types of models with the same sentences read by human subjects. RSA was used to evaluate the correlation between model embeddings and brain activation.
Credit
© 2024 Research and Innovation Office, The Hong Kong Polytechnic University. All Rights Reserved.
Usage Restrictions
Nil
License
Original content