[Call for Papers] The Causal Reinforcement Learning Workshop @RLC 2025

[Davide Corsi]

Dear all,

Are you interested in understanding the difference between correlation and causation in the context of reinforcement learning? Could causal reasoning be the next step toward building more robust, generalizable, and interpretable RL agents?

We invite you to submit your work to the Causal Reinforcement Learning (CausalRL) Workshop, which will be held on August 5th, 2025, as part of the Reinforcement Learning Conference (RLC 2025).

Call for Papers:

The CausalRL Workshop brings together researchers at the intersection of Reinforcement Learning and Causal Inference to explore how these two powerful frameworks can be combined to improve decision-making under uncertainty. We welcome contributions from both academia and industry that investigate how causal principles can enhance RL, and vice versa.

Topics of interest include (but are not limited to):

• Causal representation learning for RL

• Causal discovery and structural learning in interactive environments (MDPs, POMDPs, or SCMs)

• Counterfactual reasoning for policy/value learning

• Offline RL with unobserved confounders

• Generalization, robustness, and safety via causal reasoning

• Multi-agent causal reinforcement learning

Confirmed Speakers:

• Elias Bareinboim, Columbia University
  
• Hengrui Cai, University of California - Irvine
  
• Biwei Huang, University of California - San Diego
  
• Nathan Kallus, Cornell University
  
• Peter Stone, University of Texas at Austin and Sony AI
  
• Junzhe Zhang, Syracuse University
  
• Kun Zhang, Carnegie Mellon University

Important Dates:

• Submission Deadline: June 6th, 2025 (AOE)

• Notification of Acceptance: June 30th, 2025

  
  

Submission Details:

We welcome submissions of papers between 4 and 8 content pages in length (excluding references and appendices). Submissions may include theoretical results, empirical findings, position papers, or negative results that provoke thoughtful discussion. We especially encourage early-stage work and cross-disciplinary perspectives.

For more information, please visit our website.

Contact:

Feel free to reach out to us at dcorsi[at]uci.edu or ml[at]cs.columbia.edu with any questions. Looking forward to your submissions and to insightful discussions at the workshop!

With best regards,

Davide Corsi and Mingxuan Li, on behalf of the CausalRL Workshop organizers