{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/covariant-compositional-networks-for-learning","title":"Covariant Compositional Networks For Learning Graphs","arxiv_id":"1801.02144","date":"2018-01-07","proceeding":"ICLR 2018 1","authors":["Risi Kondor","Hy Truong Son","Horace Pan","Brandon Anderson","Shubhendu Trivedi"],"abstract":"Most existing neural networks for learning graphs address permutation\ninvariance by conceiving of the network as a message passing scheme, where each\nnode sums the feature vectors coming from its neighbors. We argue that this\nimposes a limitation on their representation power, and instead propose a new\ngeneral architecture for representing objects consisting of a hierarchy of\nparts, which we call Covariant Compositional Networks (CCNs). Here, covariance\nmeans that the activation of each neuron must transform in a specific way under\npermutations, similarly to steerability in CNNs. We achieve covariance by\nmaking each activation transform according to a tensor representation of the\npermutation group, and derive the corresponding tensor aggregation rules that\neach neuron must implement. Experiments show that CCNs can outperform competing\nmethods on standard graph learning benchmarks.","url_abs":"http://arxiv.org/abs/1801.02144v1","url_pdf":"http://arxiv.org/pdf/1801.02144v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"covariant-compositional-networks-for-learning","repo_url":"https://github.com/HyTruongSon/GraphFlow","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"none","reach":{"status":"unanswered"}},{"paper_slug":"covariant-compositional-networks-for-learning","repo_url":"https://github.com/HyTruongSon/InvariantGraphNetworks-Pytorch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[{"task_slug":"graph-learning","task_name":"Graph Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1801.02144","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}