DIA-HARM evaluates 16 harmful content detection models across 50 English dialects using 195K+ samples, revealing 1.4–3.6% F1 drops for fine-tuned models and up to 27% for zero-shot …
Area Chair (OpenReview) for the 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026), San Diego, California.
Program Committee Member for LoResLM 2026, the Workshop on Low-Resource Languages and Multilingual NLP, co-located with EACL in Rabat, Morocco.
Organizing Committee Member for the 12th Mid-Atlantic Student Colloquium on Speech, Language & Learning (MASC-SLL) at Penn State.
In this talk, we present Beemo, one of the first multi-author benchmarks for machine-generated text (MGT) detection that includes expert-edited responses. Our benchmark comprises …
Served as reviewer for top-tier NLP conferences including ACL, EACL, and EMNLP (2024–2025).
The Dual Role of LLMs in Crafting and Detecting Elusive Disinformation
Volunteered at EMNLP 2023, providing technical support, poster session coordination, and participant registration assistance.