{"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/equivariance-through-parameter-sharing","title":"Equivariance Through Parameter-Sharing","arxiv_id":"1702.08389","date":"2017-02-27","proceeding":"ICML 2017 8","authors":["Siamak Ravanbakhsh","Jeff Schneider","Barnabas Poczos"],"abstract":"We propose to study equivariance in deep neural networks through parameter\nsymmetries. In particular, given a group $\\mathcal{G}$ that acts discretely on\nthe input and output of a standard neural network layer $\\phi_{W}: \\Re^{M} \\to\n\\Re^{N}$, we show that $\\phi_{W}$ is equivariant with respect to\n$\\mathcal{G}$-action iff $\\mathcal{G}$ explains the symmetries of the network\nparameters $W$. Inspired by this observation, we then propose two\nparameter-sharing schemes to induce the desirable symmetry on $W$. Our\nprocedures for tying the parameters achieve $\\mathcal{G}$-equivariance and,\nunder some conditions on the action of $\\mathcal{G}$, they guarantee\nsensitivity to all other permutation groups outside $\\mathcal{G}$.","url_abs":"http://arxiv.org/abs/1702.08389v2","url_pdf":"http://arxiv.org/pdf/1702.08389v2.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":"equivariance-through-parameter-sharing","repo_url":"https://github.com/raymondyeh07/chirality_nets","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"sensitivity","task_name":"Sensitivity"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1702.08389","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}