Max Simchowitz

Max Simchowitz msimchow(AT)andrew(DOT)cmu(DOT)edu CV Google Scholar profile

About Me

I am an incoming assistant professor in the Machine Learning Department at Carnegie Mellon University, beginning in January 2025. I will be spending this Fall as a visiting researcher at the Modern Paradigms in Generalization program at the Simons Institute for the Theory of Computing, at UC Berkeley. Moreover: I am actively recruiting students applying for PhD programs this year to begin in Fall of 2025!

My current work focuses on the theoretical foundations of learning for robotics and how to operationalize these insights in practical algorithms; this research has motivated a parallel interest in learning and extrapolation under distribution shift. My current work draws on past research ranging broadly across topics in adaptive sampling, multi-arm bandits, complexity of convex and non-convex optimization, reinforcement learning, learning in linear and nonlinear dynamical systems, and fairness in machine learning.

I was formerly a postdoc in Russ Tedrake's Robot Locomotion Group in the EECS Department at MIT. I received my PhD in the EECS department at UC Berkeley, co-advised by Ben Recht and Michael Jordan, where I was generously supported by Open Philanthropy, NSF GRFP grant and Berkeley Fellowship grants. I also worked with, and was closely mentored by, Elad Hazan at Princeton and Kevin Jamieson, now at University of Washington. Previously, I recieved a BA in Mathematics at Princeton University, where I was fortunate enough to do research with Sanjeev Arora and David Blei (who taught at Princeton at the time).

Selected Works

• Diffusion Forcing: Next-token Prediction Meets Full-Sequence Diffusion
  Boyuan Chen, Diego Marti Monso, Yilun Du, Max Simchowitz, Russ Tedrake, Vincent Sitzmann. Under Submission, 2024.

• Provable Guarantees for Generative Behavior Cloning: Bridging Low-Level Stability and High-Level Behavior
  Adam Block*, Ali Jadbabaie, Daniel Pfrommer*, Max Simchowitz*, Russ Tedrake. Neurips, 2023. (* denotes equal contribution)
  Former title, currently used by Google Scholar: "Imitating Complex Behavior: Bridging Low-Level Stability and High-Level Behavior"

• Statistical Learning under Heterogeneous Distribution Shift
  Max Simchowitz, Anurag Ajay, Pulkit Agrawal, Akshay Krishnamurthy. ICML, 2023.

• Do Differentiable Simulators Give Better Policy Gradients?
  H.J. Terry Suh, Max Simchowitz, Kaiqing Zhang, Russ Tedrake. ICML, Outstanding Paper Award, 2022.

• Naive Exploration is Optimal for Online LQR
  Max Simchowitz, Dylan Foster. ICML, 2020.

• Improper Learning for Nonstochastic Control
  Max Simchowitz, Karan Singh, Elad Hazan. COLT, 2020.

• Non-Asymptotic Gap-Dependent Regret Bounds for Tabular MDPs
  Max Simchowitz, Kevin Jamieson. NeurIPS, 2019.

• Learning Without Mixing: Towards A Sharp Analysis of Linear System Identification
  Max Simchowitz, Horia Mania, Stephen Tu, Benjamin Recht, Michael I. Jordan. COLT, 2018.
  

Teaching

CS 189/289A, Introduction to Machine Learning, UC Berkeley Fall 2018 (TA).

EE227C, Convex Optimization and Approximation, UC Berkeley, Spring 2018 (TA). Link for course notes.