Aligning Reinforcement Learning Experimentalists and Theorists
Workshop at the International Conference on Machine Learning, Friday, the 26 of July 2024. Vienna, Austria.
Recent progress in reinforcement learning (RL) has powered breakthroughs in various real-world problems (e.g., 1, 2, 3, 4, 5, 6), gathering considerable attention and investment. However, it has also exposed a significant gap between theoretical and experimental developments.
RL theory has grown significantly in the past two decades. Research has characterized the inherent difficulty of various settings, and has designed a wide variety of algorithms (e.g., 7, 8, 9) to reach optimal performances. Furthermore, a huge leap has been made in understanding how to handle large state spaces using function approximation techniques, identifying key structural properties that enable efficient learning (e.g., 10, 11, 12).
However, despite theoretical guarantees, applying RL algorithms to complex problems faces challenges. Theoretical algorithms often focus on simplified settings, making them hard to apply to real-world complexities. Furthermore, optimizing for worst-case scenarios, which include unlikely situations, can lead to algorithms that perform poorly on practical tasks. Yet, while specialized algorithms offer empirical success, they might not translate to other problems due to their specific design, and the reliance on heuristics and engineering fixes (13) further widens the gap between theory and practice.
With this workshop, we aim to bring theorists and experimentalists together to drive future research in RL.
RL theory has grown significantly in the past two decades. Research has characterized the inherent difficulty of various settings, and has designed a wide variety of algorithms (e.g., 7, 8, 9) to reach optimal performances. Furthermore, a huge leap has been made in understanding how to handle large state spaces using function approximation techniques, identifying key structural properties that enable efficient learning (e.g., 10, 11, 12).
However, despite theoretical guarantees, applying RL algorithms to complex problems faces challenges. Theoretical algorithms often focus on simplified settings, making them hard to apply to real-world complexities. Furthermore, optimizing for worst-case scenarios, which include unlikely situations, can lead to algorithms that perform poorly on practical tasks. Yet, while specialized algorithms offer empirical success, they might not translate to other problems due to their specific design, and the reliance on heuristics and engineering fixes (13) further widens the gap between theory and practice.
With this workshop, we aim to bring theorists and experimentalists together to drive future research in RL.
Highlights
A panel to discuss the current state of RL
Idea track: send your problems
Research track: send your solutions
We are excited to have an idea track along with the canonical call for papers, which we hope will foster increased collaboration within the community. The talks will also be followed by a panel dicussion on the current state of RL. See the schedule for details.
Speakers
We are thrilled to have the following researchers joining us for the event.
Organizers
Sponsors
Thanks to our sponsors for supporting the workshop.
Contact
You can reach us by email at arlet.icml2024@gmail.com
, or on Twitter, @arlet_workshop
.