Call for Papers

We welcome any contributions on ongoing research at the interface of causal abstractions, causal representations learning, multi-scale causal systems, including but not limited to: 

• Causal Abstraction

• Cluster and Multivariate DAGs

• Causal Representation Learning

In particular,  we invite works that define novel  frameworks, tasks, methodologies, and real-world applications, to name a few. Further, we welcome position papers on related arguments, such as:

• critical perspectives on the pragmatic usability of causal abstraction and representational learning, 

• connections between causal abstraction and representational learning and multivariate causal inference,

• development of benchmarks and evaluation metrics

We invite submissions of ongoing research that has not yet been published in a venue with proceedings. While we welcome short papers and works-in-progress, submissions should present original ideas, novel connections between research areas, or preliminary results with potential impact.

The main body of the submission—including figures and tables must not exceed 8 pages. Supplementary material may be included in the same PDF, placed after the references, and is not subject to a page limit. However, it is entirely at the reviewers’ discretion whether they choose to consult this additional content.

Accepted papers will be allowed one additional page for the camera-ready version, for a total of up to 9 pages.

We also encourage authors to include links to public repositories (e.g. GitHub) to share code and/or data.

Papers must be submitted anonymously as a single PDF to the OpenReview submission portal and formatted according to the workshop’s style file. Submissions that do not comply with these requirements will be automatically rejected.

All papers will undergo double-blind peer review. A subset of accepted papers will be selected for contributed talks; the remainder will be presented in the poster sessions. Please note that this workshop will not have formal proceedings.