2157 N Northlake Way
Seattle, WA, USA 98103
My research interests include studying bias in datasets - the good, the bad and the ugly. Good biases, such as structural inductive biases beneficial for language understanding, I wrote a PhD thesis on these. On the other hand, crowd-sourced, large-scale datasets are riddled with annotation artifacts which are spurious correlations with unintended effects; I call these the bad biases. And finally, biases can be ugly, when the training data contains a large portion of mislabeled examples, or noise (more on this soon!).
I obtained my PhD from Carnegie Mellon University in May 2019, where I was advised by Noah Smith and Chris Dyer. During most of my PhD I was a visiting student at the Paul G. Allen School of Computer Science at the University of Washington in Seattle.
|Apr 11, 2020||New EMNLP preprint on Generative Data Augmentation for Commonsense Reasoning, or G-DAUG now available on arXiv.|
|Apr 3, 2020||Two papers accepted to ACL! Kudos to my wonderful collaborators on Don’t Stop Pretraining: Adapt Language Models to Domains and Tasks and The Right Tool for the Job: Matching Model and Instance Complexities.|
|Feb 2, 2020||New preprint on Adversarial Filters of Dataset Biases now available on arXiv.|
|Oct 17, 2019||Invited talk at the UW Linguistics Colloquium on Oct 18, 2019.|
|Oct 2, 2019||Got selected to attend the Rising Stars in EECS workshop at the University of Illinois at Urbana-Champaign from Oct 29 - Nov 1, 2019.|