2026
PhD student · UT Austin
Max Rudolph
I work on deep reinforcement learning — unsupervised structure in decision-making, and what that buys us on real robots.
About
Research, in short
I am a PhD student at UT Austin advised by Amy Zhang, working on deep reinforcement learning. I am interested in a wide range of RL problems; lately that means unsupervised learning in decision-making, and applying those tools to large-scale robot learning.
Previously I worked in the STAR lab with Harish Ravichandar and the RAIL lab with Sonia Chernova at Georgia Tech. I also collaborated closely with Joanne Truong on sim-to-real for legged robots.
Selected work
Papers worth starting with
2025
2024
2023
2022
Rethinking Sim2Real: Lower Fidelity Simulation Leads to Higher Sim2Real Transfer in Navigation
Updates
What I’ve been up to
- Paper on reevaluating policy gradient methods for imperfect-information games accepted at ICLR 2026.
- RL Zero accepted at NeurIPS 2025.
- Oral presentation at the RLC 2025 CoCoMARL workshop.
- Workshop on building trust in AI systems accepted at Trustworthy Autonomous Systems 2024.
- Paper on bisimulation representation learning accepted at the Reinforcement Learning Conference.
- Started an Applied Scientist internship at Amazon on exploration for reinforcement learning.
- Named a Finalist for the Qualcomm Innovation Fellowship.
- Awarded the NSF NRT Ethical AI Fellowship.
- Started a CS PhD at UT Austin, advised by Amy Zhang.
Also
Off the clock, and a few other things
I play tennis and make bean-to-bar chocolate and bon bons. I run the CS graduate student organization GRACS at UTCS.
During undergrad I interned at the NSA and at NASA Jet Propulsion Lab, working on software for the Mars Perseverance Rover and control algorithms for the Psyche mission.
