{"ID":2846275,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2511.02667","arxiv_id":"2511.02667","title":"Scalable Evaluation and Neural Models for Compositional Generalization","abstract":"Compositional generalization-a key open challenge in modern machine learning-requires models to predict unknown combinations of known concepts. However, assessing compositional generalization remains a fundamental challenge due to the lack of standardized evaluation protocols and the limitations of current benchmarks, which often favor efficiency over rigor. At the same time, general-purpose vision architectures lack the necessary inductive biases, and existing approaches to endow them compromise scalability. As a remedy, this paper introduces: 1) a rigorous evaluation framework that unifies and extends previous approaches while reducing computational requirements from combinatorial to constant; 2) an extensive and modern evaluation on the status of compositional generalization in supervised vision backbones, training more than 5000 models; 3) Attribute Invariant Networks, a class of models establishing a new Pareto frontier in compositional generalization, achieving a 23.43% accuracy improvement over baselines while reducing parameter overhead from 600% to 16% compared to fully disentangled counterparts. Our code is available at https://github.com/IBM/scalable-compositional-generalization.","short_abstract":"Compositional generalization-a key open challenge in modern machine learning-requires models to predict unknown combinations of known concepts. However, assessing compositional generalization remains a fundamental challenge due to the lack of standardized evaluation protocols and the limitations of current benchmarks,...","url_abs":"https://arxiv.org/abs/2511.02667","url_pdf":"https://arxiv.org/pdf/2511.02667v2","authors":"[\"Giacomo Camposampiero\",\"Pietro Barbiero\",\"Michael Hersche\",\"Roger Wattenhofer\",\"Abbas Rahimi\"]","published":"2025-11-04T15:45:45Z","proceeding":"cs.LG","tasks":"[\"cs.LG\",\"cs.AI\"]","methods":"[]","has_code":false,"code_links":[{"ID":607426,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_id":2846275,"paper_url":"https://arxiv.org/abs/2511.02667","paper_title":"Scalable Evaluation and Neural Models for Compositional Generalization","repo_url":"https://github.com/IBM/scalable-compositional-generalization","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}