Post-Doc Researcher (Saarland University)
msullivan at lst dot uni-saarland dot de
Semantics/Pragmatics Track
Thesis: Language Modeling over Logical Forms
University at Buffalo (2020-2025)
Research/Honors Track
MS Project: Probing NLI Models with External Negation
University at Buffalo (2023-2024)
With Research Distinction
Minors: Spanish, German
The Ohio State University (2016-2019)
I am a Postdoc in the Computational Linguistics Group at UdS and an Associated Member of the Research Training Group "Neuroexplicit Models of Language, Vision, and Action".
My research interests lie in the use of formal/theoretic tools to improve LLM-based systems. In particular, I am interested in leveraging formal structure to make LLM training and inference more efficient: for example, detecting equivalence-class structures to control for training-data irregularities in reinforcement learning (λ-GRPO), exploiting redundancies in the token vocabulary to speed up constrained decoding (CFGZip), or using type theory to generate verifiably coherent tool-use training data (RandomWorld).
My PhD is in Linguistics (under JP Koenig) and my MSc is in Computer Science and Engineering (under Rohini K Srihari)—both at the University at Buffalo. My Erdős number is four.