RL as a Model of Agency

Perspectives, Limitations, & Possibilities

@ RLDM 2022

13:00 - 17:00 in Macmillan Hall Room 117

Saturday, June 11, 2022


Is the RL problem a suitable model of agency?

In this workshop we bring together researchers at the intersection of artificial intelligence, cognitive science, and philosophy to carefully examine the reinforcement learning paradigm, the assumptions that underpin it, and its relationship with the broader concept of agency. We believe now is a critical time to improve clarity on our foundations, and we believe this workshop can help expose & explore new approaches, limitations, and perspectives across learning & agency. Our primary goal is that all attendees leave with at least one new question they had not thought about before (and perhaps new collaborators to help think through it!)


Professor of Philosophy at U. Toronto

Research Scientist at DeepMind

Author of "The Alignment Problem" & "Algorithms to Live By"

Professor of Computer Science at UIUC

Call for Papers

Submission Deadline: May 4th, 2022, anywhere on Earth.

Acceptance Notification: Around May 11th, 2022.

Camera Ready Due: June 1st, 2022, anywhere on Earth. Please email the version you would like posted to the website to rlasagency@gmail.com.

Topics. We invite submission of papers of 1-3 content pages (excluding the title page, references, acknowledgements, and appendices) that draw attention to a single question or perspective on any of the following topics:

  • Agency, embodiment, time, physics, and the environment.

  • Agency, goals, reward-maximization, value, and alignment.

  • Agency, abstraction, representation, and introspection.

  • Agency, theory of mind, game theory, communication, and social welfare.

Submission Instructions. Submissions should be emailed to rlasagency@gmail.com as a pdf, formatted using the RLDM author kit (latex sty file, rtf). All submissions will be reviewed single-blind, with accepted abstracts posted on the workshop website. Time-depending, selected abstracts will be presented at the workshop in a short talk.