Wes Gurnee

Wes Gurnee

PhD Student



I am a third year PhD student in Operations Research at the MIT ORC advised by Dimitris Bertsimas studying large-scale optimization, broadly applied. I am especially interested in understanding and reducing the risks posed by advanced artificial intelligence systems. Currently, this involves deeply understanding the learned representations and circuits of LLMs in the hope of maintaining oversight and diagnosing failures of increasingly capable models. In the past, I have done research on dynamical systems modeling, fair algorithms for congressional redistricting, and natural language processing.


  • Mechanistic Interpretability
  • Optimization
  • AI Safety
  • Governance


  • PhD in Operations Research, 2025

    Massachusetts Institute of Technology

  • BS in Computer Science, 2020

    Cornell University


Quickly discover relevant content by filtering publications.
(2023). Finding Neurons in a Haystack: Case Studies with Sparse Probing. Preprint.

Preprint Code

(2022). Learning Sparse Nonlinear Dynamics via Mixed-Integer Optimization. Nonlinear Dynamics.

Preprint Code

(2021). Combatting Gerrymandering with Social Choice: the Design of Multi-member Districts. Under Review.




Software Engineer


Aug 2020 – Aug 2021
Engineer on the Storage Efficiency team. Working on building a canonical source for storage fleet telemetry data.

Executive Director


Jun 2020 – Present
Founder and executive director of Fairmandering, an organization focused on data driven redistricting reforms.