Published a paper at EACL 2023 about how prompting for explanations improves adversarial NLI; presented COPA-SSE: Semi-structured Explanations for Commonsense Reasoning at the Thirteenth Language Resources and Evaluation Conference; published a study on whether prompt-based models are clueless at ACL 2022.
Research Experience
Worked on the scue project, a unified framework for analyzing superficial cues in datasets.
Education
PhD in NLP from Tohoku University, Japan in 2023; Master of Information Science (NLP) from Tohoku University, Japan in 2020; BEng in Electrical Engineering and Electronics Engineering from The University of Zambia in 2013.
Background
Research interests include LLMs, evaluation, robustness, interpretability, and common sense reasoning.