معرفی
Mohit Iyyer is an Associate Professor in Computer Science at the University of Maryland, College Park (UMD), affiliated with the Manning College of Information and Computer Sciences (CICS) and the CLIP lab. Previously, he held roles at UMass Amherst, AI2, and completed his PhD at the University of Maryland. His research focuses on advancing NLP and machine learning, particularly in long-context LLMs, factuality evaluation, and human-LLM collaboration.
Education:
- PhD, Computer Science, University of Maryland, College Park (2017)
- MS, Computer Science, University of Maryland, College Park (2014)
- BS, Computer Science, Washington University (2012)
Research Interests:
His work emphasizes improving instruction-following in LLMs, evaluating long-form text, detecting AI-generated content, and enhancing LLM robustness. He actively explores multilingual LLMs, literary translation, and creative writing support through AI.
Key Contributions:
Developed influential frameworks like ELMo and QuAC, and introduced benchmarks such as BEARCUBS and BooookScore. His work on watermarking (PostMark) and adversarial attacks (OverThink) addresses security and reliability in AI systems.
Awards & Recognition:
- NSF CAREER Award (2021)
- Samsung AI Researcher of the Year (2022)
- Best Paper Awards at NAACL 2018 and CCS 2023
Advising & Grants:
Advises a dynamic research group and has received grants for projects like multilingual QA and long-context LLM evaluation. Collaborates with industry partners like AI2 and DeepMind.
Labs & Teams:
Leads the CLIP lab at UMD, focusing on foundational NLP research and applications in literature, dialogue systems, and computational storytelling.
