I am a PhD student in the Cognitive Science department at Johns Hopkins University working with Tal Linzen . I am interested how humans and neural language models represent sentences – specifically, what information about structure and statistical regularities is encoded in their sentence representations. I am also interested in how these representations change with experience. Working on these questions involves running psycholinguistic experiments on humans/ language models and thinking about how to compare the human representation space with that of the language models.

Apart from this, I care about open science, replicability and issues of statistical power. I would also almost always be up for discussions about philosophy of (cognitive) science, or any philosophical discussion really.

Outside of work, I get immense satisfaction from drawing parallels and making connections between seemingly obscure things - whether it is for some really convoluted pun or to draw analogies from connectionist theory when thinking about relationships. I also occasionally write haikus describing my state of mind - I find that the structural restrictions can sometimes help synthesize my thoughts (much like how a neural network might learn better when it has to compress the information from the input in a few units).