This call for workshop papers overwrites the main CfP. Please read it first. Here, we only highlight how this workshop track differs from the main conference track. For dates, please see here.


Topics include all topics from the main CfP.


As the name suggests, this workshop track provides a venue for sharing your latest work in a workshop format, much like the AutoML workshop series at ICML used to. By its very nature, this latest work may not be as comprehensive yet as full conference papers, and correspondingly, submissions to this workshop track are short papers. From our side, these short workshop papers without formal proceedings are fully compatible with the later publication of an extended version at a conference, just like is the case for most workshops at NeurIPS, ICML and ICLR. We explicitly encourage the submission of short versions of work that is in preparation for, under review or accepted at top conferences occurring within 6 months of AutoML 2024 (a non-exhaustive list includes ICLR’24, ICML’24, NeurIPS’24). All submissions will be reviewed, but with a faster turn-around than at the main track and without rebuttal.

Submission Guidelines

Formatting Instructions

In contrast to the main track, this track only allows short papers of up to 5 pages, including the broader impact and limitations statement, in contrast to the 9 pages at the main conference. We encourage splitting these 5 pages as 4+1 pages into 4 pages for the main paper and 1 page on broader impact and limitations (combined). References and appendices do not count into this limit and are not limited in length. Like in the methods track, the submission should be double-blind. However, if the submission falls into the scope of the ABCD track, it can be single-blind.


We strongly encourage you also to submit the reproducibility checklist and open up your source code and data. However, this is not required for this track. Nevertheless, we will give reproducibility badges to all accepted submissions with a reproducibility checklist, open-source code and open-access data.

Relation to the main track’s “accept as workshop paper” category

This year, method and ABCD track decisions are ternary: accept, accept as workshop paper, or reject. Papers in the second of these categories are accepted as is to the workshop track unless the authors opt out. In this case, the paper would remain as is and cannot be shortened to 5 pages (as that would change the paper too much). Authors of such papers could also opt out and instead submit a short version to the workshop track, but such short versions would go through the standard workshop track review and have no guarantee of acceptance to the workshop.

The PC chairs also chair this track.