Call for Late-Breaking Abstracts

We are excited to announce a call for Late-Breaking Abstracts for the 2026 AutoML Conference in Ljubljana. This track is designed for fresh ideas and ongoing projects that were not ready for the main conference deadline. The goal is to foster an interactive, productive environment where authors can discuss unpolished methods and visionary concepts with the community to kickstart future collaborations.

Submissions must be made via our Google Forms Link. Please prepare the following:

1. The Paper: A maximum of 2 pages (excluding references) using the standard AutoML LaTeX template ( https://github.com/automl-conf/LatexTemplate ).
2. The “TL;DR” abstract: A brief 2–3 sentence summary of your work. This must be included both in the paper as the abstract and as a text entry in the Google Form.
3. Discussion questions: Provide 3 open questions or challenges related to your work that you wish to discuss with attendees during your poster session.

We welcome all topics within the scope of the  main tracks (Methods and ABCD). Because this track emphasizes fresh ideas over polished benchmarks, we also encourage non-traditional content, such as new open-source tools or software prototypes, blog-style write-ups, or position papers on the future of AutoML. A note on scope: While we welcome tools and blog-style insights, ensure the core focus remains relevant to Automated Machine Learning. Purely general ML tools without an automation component may be considered out of scope.

Accepted abstracts will be presented during a dedicated poster session at the conference.

• Poster printing: Authors are responsible for printing and bringing their physical posters to the venue in Ljubljana.
• Interactive format: The 3 open questions provided during submission will be used to facilitate deep-dive discussions at your poster. Ideally, you also have them on your poster. This ensures you leave the session with actionable feedback.
• Registration: At least one author must register for and attend the conference in person.

This is the perfect venue to get eyes on your latest work-in-progress without the pressure of a full archival review. Whether you have a new library, a wild idea for hyperparameter optimization, or an interesting observation about LLM agents in the context of AutoML, we want to see it! Submit your late-breaking work by July 15th!