Home

Mridul Mahajan
mridulm [at] bu [dot] edu

Hi! I am a second-year PhD student at Boston University, advised by Aldo Pacchiano. I am broadly interested in reinforcement learning (RL) from an algorithmic perspective, with a focus on how transfer, structure, and diversity can improve upon tabula-rasa learning in terms of sample efficiency and performance.

Previously, I was a guest researcher in the Empirical Inference department at the Max Planck Institute for Intelligent Systems (MPI-IS), and a visiting scholar in the Machine Teaching Group at the Max Planck Institute for Software Systems (MPI-SWS).

email   |   cv   |   google scholar   |   github

profile photo
Recent Publications

* indicates equal contribution and co-authorship.

Reinforcement Learning

PolyNet: Learning Diverse Solution Strategies for Neural Combinatorial Optimization
A. Hottung, M. Mahajan, K. Tierney
International Conference on Learning Representations (ICLR) 2025
paper

Learning Embeddings for Sequential Tasks Using Population of Agents
M. Mahajan, G. Tzannetos, G. Radanovic, A. Singla
International Joint Conference on Artificial Intelligence (IJCAI) 2024
paper

Robotics

Sim-to-Real Transfer for Muscle-Actuated Robots via Generalized Actuator Networks
J. Schneider, M. Mahajan, L. Chen, S. Guist, B. Schölkopf, I. Posner, D. Büchler
Preprint 2026
paper

METEOR: A Dense, Heterogeneous, and Unstructured Traffic Dataset With Rare Behaviors
R. Chandra*, X. Wang*, M. Mahajan, R. Kala, R. Palugulla, C. Naidu, A. Jain, D. Manocha
IEEE International Conference on Robotics and Automation (ICRA) 2023
paper



website adapted from here