Pablo Samuel Castro

Researcher, Google

Joined

September 2018

Names

Pablo Samuel Castro (Preferred)

Emails

****@google.com (Confirmed)

****@mila.quebec (Confirmed)

****@umontreal.ca (Confirmed)

Personal Links

Homepage

Google Scholar

DBLP

Career & Education History

Researcher

Google (google.com)

2012 – Present

PhD student

McGill University (mcgill.ca)

2007 – 2011

MS student

McGill University (mcgill.ca)

2005 – 2007

Advisors, Relations & Conflicts

PhD Advisor

Doina Precup

2005 – 2011

PhD Advisor

Prakash Panangaden

2005 – 2011

Expertise

Reinforcement Learning

2005 – Present

Publications

Discovering Differences in Strategic Behavior between Humans and LLMs
Caroline Wang, Daniel Kasenberg, Kim Stachenfeld, Pablo Samuel Castro
- HCAIR 2026
- Readers: Everyone
ARM-FM: Automated Reward Machines via Foundation Models for Compositional Reinforcement Learning
Roger Creus Castanyer, Faisal Mohamed, Pablo Samuel Castro, Cyrus Neary, Glen Berseth
- ICLR 2026 Poster
- Readers: Everyone
Simplicial Embeddings Improve Sample Efficiency in Actor–Critic Agents
Johan Obando-Ceron, Walter Mayor, Samuel Lavoie, Scott Fujimoto, Aaron Courville, Pablo Samuel Castro
- ICLR 2026 Poster
- Readers: Everyone
Asymmetric Proximal Policy Optimization: mini-critics boost LLM reasoning
Jiashun Liu, Johan Obando-Ceron, Han Lu, Yancheng He, Weixun Wang, Wenbo Su, Bo Zheng, Pablo Samuel Castro, Aaron Courville, Ling Pan
- ICLR 2026 Poster
- Readers: Everyone
A Comedy of Estimators: On KL Regularization in RL Training of LLMs
Vedant Shah, Johan Obando-Ceron, Vineet Jain, Brian R. Bartoldson, Bhavya Kailkhura, Sarthak Mittal, Glen Berseth, Pablo Samuel Castro, Yoshua Bengio, Nikolay Malkin, Moksh Jain, Siddarth Venkatraman, Aaron Courville
- Submitted to ICLR 2026
- Readers: Everyone
Align and Filter: Improving Performance in Asynchronous On-Policy RL
Homayoun Honari, Roger Creus Castanyer, Michael Przystupa, Michael Noukhovitch, Pablo Samuel Castro, Glen Berseth
- Submitted to ICLR 2026
- Readers: Everyone
What Matters for Maximizing Data Reuse In Value-based Deep Reinforcement Learning
Roger Creus Castanyer, Glen Berseth, Pablo Samuel Castro
- Finding the Frame (RLC 2025)
- Readers: Everyone
Training PPO-Clip with Parallelized Data Generation: A Case of Fixed-Point Convergence
Homayoun Honari, Roger Creus Castanyer, Pablo Samuel Castro, Glen Berseth
- IBRL @ RLC 2025
- Readers: Everyone
A Geometric Lens on RL Environment Complexity Based on Ricci Curvature
Ali Saheb Pasand, Pablo Samuel Castro, Pouya Bashivan
- RLBrew: Ingredients for Developing Generalist Agents workshop (RLC 2025)
- Readers: Everyone
A Geometric Lens on RL Environment Complexity Based on Ricci Curvature
Ali Saheb Pasand, Pablo Samuel Castro, Pouya Bashivan
- Finding the Frame (RLC 2025)
- Readers: Everyone

View all 112 publications

Co-Authors

View all 203 co-authors