Sean Welleck

Sean Welleck is an Assistant Professor at Carnegie Mellon University's School of Computer Science, Language Technologies Institute, leading the L3 Lab. His research focuses on bridging informal and formal reasoning with AI, spanning machine learning for mathematics and code, inference algorithms, and AI agents.

• PhD in Computer Science from New York University (advised by Kyunghyun Cho)
• Postdoctoral work at University of Washington (advised by Yejin Choi)

His work explores AI-driven formal methods for mathematics and code generation, test-time compute scaling, and algorithms enabling AI improvement over time. Recent publications analyze reasoning evaluation, premise selection, and automated proof optimization in systems like Lean.

Key article trends include neural theorem proving, code generation, and inference-time compute optimization. Awards: NVIDIA AI Labs Pioneering Research Awards (2017, 2018), NAACL 2025 Best Paper.

Current advisees include PhD students Pranjal Aggarwal, Weihua Du (co-advised with Yiming Yang), Andre He (co-advised with Daniel Fried), and Seungone Kim (co-advised with Graham Neubig).

He co-organizes workshops like Autoformalization for the Working Mathematician (ICERM 2025) and VerifAI: AI Verification in the Wild (ICLR 2025), and teaches Advanced NLP at CMU.