I am a Ph.D. student in Computer Science at the University of Massachusetts Lowell, where I work in the Computational Language Understanding (CLU) Lab advised by Professor Hadi Amiri. Previously, I received my BS in Computer Science from KAIST.

My research aims to develop reliable and efficient NLP systems through understanding of model behavior:

• Controlled Generation: I develop techniques that enable precise steering of language model outputs.
• Curriculum Learning: I investigate data selection and ordering strategies to advance how we train large language models, aiming to enable next-generation models at a fraction of the current computational cost.
• Interpretability: I analyze model behavior to understand how models learn and steer their outputs.

Selected Publications

Full publication list →

• MedDec: A Dataset for Extracting Medical Decisions from Clinical Narratives
  Mohamed Elgaar, Jiali Cheng, Nidhi Vakil, Hadi Amiri, Leo Anthony Celi
  Findings of ACL 2024 [PDF] [Code]

• Ling-CL: Multiview Curriculum Learning using Linguistic Complexity
  Mohamed Elgaar, Hadi Amiri
  EMNLP 2023 [PDF] [Code]

• HuCurl: Human-induced Curriculum Discovery
  Mohamed Elgaar, Hadi Amiri
  ACL 2023 [PDF] [Code]

News

• [Oct 2024] Released MedDecXtract demo
• [Aug 2024] Released LingConv demo
• [Aug 2024] Presented MedDec at ACL 2024
• [June 2024] Papers accepted to Findings of ACL and INTERSPEECH 2024
• [Dec 2023] Presented Ling-CL at EMNLP 2023
• [Nov 2023] Paper accepted to EMNLP 2023
• [July 2023] Presented HuCurl at ACL’23 in Toronto
• [May 2023] Paper accepted to ACL 2023