Published multiple papers at top NLP and AI conferences including ACL, EMNLP, NAACL, CoNLL, NeurIPS, and COLING
2021: Research on negation-awareness, dynamic forecasting of conversation derailment, and implicit causality bias in language models
2020: Created the ApposCorpus, a multilingual multi-domain dataset for factual appositive generation; contributed to PuzzLing Machines challenge on learning from small data
2019: Investigated misleading benchmarks in bilingual dictionary induction, effects of grammatical gender on noun representations, low-resource dependency parsing, adversarial removal of demographic attributes, and probabilistic generative models of linguistic typology
2018: Improved Procrustes analysis for bilingual dictionary induction; proposed a discriminative latent-variable model for lexicon induction; participated in multilingual morphological inflection tasks; studied morpho-syntactic learning in character-level language models
2016: Analyzed strategic code-switching in political discourse; contributed to research on San Francisco English and the California Vowel Shift