Sean Welleck

Assistant Professor, Language Technologies Institute

Sean Welleck is an Assistant Professor at Carnegie Mellon University's Language Technologies Institute (LTI). His research focuses on the development of learning, inference, and evaluation algorithms for large language models, with a particular emphasis on applications in mathematical reasoning and code generation. Dr. Welleck has made significant contributions to areas such as inference algorithms for language models and formal theorem proving, with research featured in leading conferences like NeurIPS, ICLR, and EMNLP. Recent projects include Llemma, an open-source language model designed for mathematical reasoning, and meta-generation, a framework for inference-time algorithms for large language models.

Dr. Welleck earned his Ph.D. in Computer Science from New York University, where he worked under the mentorship of Kyunghyun Cho and Zheng Zhang, focusing on neural text generation. Before joining CMU, he was a Postdoctoral Young Investigator at the Allen Institute for Artificial Intelligence and a Postdoctoral Scholar at the University of Washington, collaborating with Yejin Choi. His research has earned numerous accolades, including best paper awards and keynote presentations, and he has co-organized workshops and tutorials on neural theorem proving at top conferences such as NeurIPS.

At CMU, Dr. Welleck teaches courses like Neural Code Generation and Advanced NLP and actively mentors students in exploring innovative AI algorithms and applications of AI in mathematics and beyond. He also serves as an advisor for the AI for Math Fund. His work continues to push the boundaries of AI's capabilities in formal reasoning and generative modeling.

Research Areas