Professor Suhas Arehalli

This research is at the intersection of Natural Language Processing and the psychology of language. The researchers train and evaluate neural network models (typically language models or neural parsers) on their linguistic abilities, with a focus on interpretability (What strategies do models adopt in order to perform linguistic tasks?) and applications to cognitive science (Do models learn strategies that resemble those of humans? How can understanding the strategies artificial language processors learn help inform theories of human language processing?).