Research Activities

Paper accepted at ACL 2021

Determining the relative importance of words in a sentence is vital for language understanding. In this work, we compare patterns of relative importance in human language processing and gradient-based saliency in neural LMs.

CMCL shared task on eye-tracking prediction

We presented the first shared task on eye-tracking prediction that allows to compare the capabilities of machine learning approaches to model and analyze human patterns of reading. The results will be presented at the CMCL workshop.

Paper accepted at NAACL 2021

Do multilingual language models accurately predict human reading behavior? We ask this question in our new paper. Preprint available here!


Student Projects