Home | Sobhan's Blog

Hi, I'm Sobhan.

I'm doing a PhD at MIT advised by Gabriele Farina working on computational game theory. I'm interested in computational tools for finding equilibriums of large games.

Previously I did an MSc and BSc at Université de Montréal . My MSc was advised by Emma Frejinger and Pierre-Luc Bacon . I was affiliated with Mila and CIRRELT where I worked on problems adjecent to reinforcement learning (RL), regularization (entropy), and traffic modeling.

You can reach me via email at first two characters of my first and last name + @mit.edu You can also find me on github @SobhanMP where I sometimes contribute to open source software.

News

[09/2025]: I'm TA'ing Accelerated Computing (6.S894)
[02/2025]: Reevaluating Policy Gradient Methods for Imperfect-Information Games is on arXiv
[02/2025]: I'm TA'ing Nonlinear Optimization (6.7220)
[01/2025]: New website

Research

arXiv

Reevaluating policy gradient methods for imperfect-information games

Max Rudolph , Nathan Lichtlé , Sobhan Mohammadpour, Alexandre Bayen , Zico Kolter , Amy Zhang , Gabriele Farina , Eugene Vinitsky , Samuel Sokota

AISTATS 2024

Maximum entropy GFlowNets with soft Q-learning

Sobhan Mohammadpour, Emmanuel Bengio , Emma Frejinger , Pierre-Luc Bacon

ICLR 2024

Decoupling regularization from the action space

Sobhan Mohammadpour, Emma Frejinger , Pierre-Luc Bacon

arXiv

Arc travel time and path choice model estimation subsumed

Sobhan Mohammadpour, Emma Frejinger