The 43rd International Conference on Machine Learning (ICML) took place on July 6–11, 2026, in Seoul, South Korea.
We’re proud to announce that SmallData researchers took the opportunity to showcase their work in this prestigious conference.
“End-to-End Compression for Tabular Foundation Models” by our PI Josif Grabocka and our doctoral researcher Guri Zabërgja was selected as a spotlight, which represents the top 2.2% of all submissions.
Their paper introduces TACO, an end-to-end tabular compression model that compresses the training dataset in a latent space. This method addresses the overhead on training and inference time of current transformer-based tabular models when scaling into bigger dataset sizes. Their results have shown 94× speedup during inference and 98% memory savings with no significant performance degradation.
Learn more about their work here.
“SurvPFN: Towards Foundation Models for Survival Predictions” by our PI Frank Hutter, our associated PI Pascal Schlosser and doctoral researchers Samuel Böhm and Lennart Purucker was showcased at the Foundation Models for Structured Data (FMSD) Workshop at ICML 2026, of which Lennart Purucker was also an organizing member.
They introduce SurvPFN, a prior-data fitted network (PFN) for survival prediction tasks. Survival prediction must account for censored data, i.e., events of interest (such as death, failure, or cure) that are not observed during the follow-up period. Standard tabular foundation models, however, cannot handle censored data. Pretrained on millions of synthetic survival prediction tasks, SurvPFN learns survival through distributional regression that accounts for censored data. Their model shows to be competitive with classical and deep survival baselines without per-dataset fitting, a survival-specific architecture, or feature engineering.
Learn more about their work here.
We are delighted by this year’s contributions to the conference and wish our researchers every success in their future work.
Institute of Medical Biometry and Statistics,
Faculty of Medicine and Medical Center –
University of Freiburg
