Decision Awareness
in Reinforcement Learning

Submit here: https://openreview.net/group?id=ICML.cc/2022/Workshop/DARL

We invite submissions with theoretical, algorithmic, and empirical contributions on topics, including, but not limited to,

• Decision-aware objectives, end-to-end procedures, and meta-learning techniques for training and discovering components in modular RL systems
• Theoretical and empirical analyses of the interaction among multiple modules used by RL algorithms
• Studies of the performance of model-based RL agents under model misspecification
• The role of unsupervised learning and auxiliary training signals in RL and their relation to decision-awareness
• Connections with other fields such as control theory, cognitive science, and robotics

We welcome submissions in both short (up to 4 pages excluding references and appendices) and long (up to 9 pages excluding references and appendices) formats. The papers should contain new material, not already accepted for publication in other venues; papers currently under review for other venues are allowed. Submissions will be reviewed in a double-blind process, managed in OpenReview. We will release talk videos on our workshop website, along with all accepted papers (non-archival proceedings, so future publication is allowed). In case authors of an accepted paper cannot present their work in person, we will ask provide a recorded video of the duration up to 5 minutes.

Evaluation Criteria: We plan to evaluate papers based on novelty, technical soundness, clarity, and potential impact. We are happy to accept works-in-progress to encourage discussion at the workshop.

We do not enforce any hard constraint on the template to be followed, but we recommend you to employ the default ICML 2022 template.

The submission deadline is May 27, 2022, Anywhere on Earth.