Hi! I am a second-year Ph.D. student in Statistics at the University of Toronto. I am very fortunate to be advised by Prof. Wenlong Mou and Prof. Xin Bing. I am also affiliated with the Vector Institute. Prior to my Ph.D. study, I received my B.S. in Statistics from Nanjing University in 2024.

My research focuses on designing statistically grounded machine learning algorithms that are both efficient and provably reliable, bridging theory and practice through optimization and statistical modeling. I currently work on LLM post-training and inference-time methods under limited feedback.

Email: muheng [DOT] li [AT] mail [DOT] utoronto [DOT] ca

News

• Feb 2026, New preprint on test-time scaling laws and budget-efficient inference-time search.
• Feb 2025, one paper accepted by TMLR.

Publications

• Predicting and Improving Test-time Scaling Laws via Reward Tail-guided Search
  Under review, 2026
  Muheng Li, Jian Qian, Wenlong Mou
  Paper · Code

• Reheated Gradient-based Discrete Sampling for Combinatorial Optimization
  Transactions on Machine Learning Research (TMLR), 2025
  Muheng Li, Ruqi Zhang
  Paper· Code

Services

• Conference Reviewer: UAI (2024, 2025); ICML (2025); ICLR (2025)

Teaching

• STA314H1F: Statistical Methods for Machine Learning I (Fall 2025), Teaching Assistant, University of Toronto.
• STA255: Statistical Theory (Winter 2025), Teaching Assistant, University of Toronto.