{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:RFNBYO57BSWI4GR5DAXAL34FYV","short_pith_number":"pith:RFNBYO57","schema_version":"1.0","canonical_sha256":"895a1c3bbf0cac8e1a3d182e05ef85c55c1f005988ff30e3e30b6b8fb0f7ad00","source":{"kind":"arxiv","id":"2302.03655","version":2},"attestation_state":"computed","paper":{"title":"Reducing SO(3) Convolutions to SO(2) for Efficient Equivariant GNNs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["physics.chem-ph","physics.comp-ph"],"primary_cat":"cs.LG","authors_text":"C. Lawrence Zitnick, Saro Passaro","submitted_at":"2023-02-07T18:16:13Z","abstract_excerpt":"Graph neural networks that model 3D data, such as point clouds or atoms, are typically desired to be $SO(3)$ equivariant, i.e., equivariant to 3D rotations. Unfortunately equivariant convolutions, which are a fundamental operation for equivariant networks, increase significantly in computational complexity as higher-order tensors are used. In this paper, we address this issue by reducing the $SO(3)$ convolutions or tensor products to mathematically equivalent convolutions in $SO(2)$ . This is accomplished by aligning the node embeddings' primary axis with the edge vectors, which sparsifies the"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2302.03655","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-02-07T18:16:13Z","cross_cats_sorted":["physics.chem-ph","physics.comp-ph"],"title_canon_sha256":"af0a79de523df08eff2d9c48511a181d2b9f02359ca5dc65d41c9e0b336a12a5","abstract_canon_sha256":"d19be0c104581b002ed443bfa91cfb5877d6b04537e5fda03449d01387b069ff"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:20:40.563029Z","signature_b64":"4x/aGUWsW5JZj/vtD2Yle0yWwDSm79VjR7whkbbBb2gh2XdzNNKRquOyIBAGSmePpOfWuSgvqWzmOa26LZAOAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"895a1c3bbf0cac8e1a3d182e05ef85c55c1f005988ff30e3e30b6b8fb0f7ad00","last_reissued_at":"2026-07-05T06:20:40.562613Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:20:40.562613Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Reducing SO(3) Convolutions to SO(2) for Efficient Equivariant GNNs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["physics.chem-ph","physics.comp-ph"],"primary_cat":"cs.LG","authors_text":"C. Lawrence Zitnick, Saro Passaro","submitted_at":"2023-02-07T18:16:13Z","abstract_excerpt":"Graph neural networks that model 3D data, such as point clouds or atoms, are typically desired to be $SO(3)$ equivariant, i.e., equivariant to 3D rotations. Unfortunately equivariant convolutions, which are a fundamental operation for equivariant networks, increase significantly in computational complexity as higher-order tensors are used. In this paper, we address this issue by reducing the $SO(3)$ convolutions or tensor products to mathematically equivalent convolutions in $SO(2)$ . This is accomplished by aligning the node embeddings' primary axis with the edge vectors, which sparsifies the"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2302.03655","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2302.03655/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2302.03655","created_at":"2026-07-05T06:20:40.562669+00:00"},{"alias_kind":"arxiv_version","alias_value":"2302.03655v2","created_at":"2026-07-05T06:20:40.562669+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2302.03655","created_at":"2026-07-05T06:20:40.562669+00:00"},{"alias_kind":"pith_short_12","alias_value":"RFNBYO57BSWI","created_at":"2026-07-05T06:20:40.562669+00:00"},{"alias_kind":"pith_short_16","alias_value":"RFNBYO57BSWI4GR5","created_at":"2026-07-05T06:20:40.562669+00:00"},{"alias_kind":"pith_short_8","alias_value":"RFNBYO57","created_at":"2026-07-05T06:20:40.562669+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.21483","citing_title":"Velocityformer: Broken-Symmetry-Matched Equivariant Graph Transformers for Cosmological Velocity Reconstruction","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15073","citing_title":"Fast contracted Clebsch--Gordan tensor products for equivariant graph neural networks","ref_index":37,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/RFNBYO57BSWI4GR5DAXAL34FYV","json":"https://pith.science/pith/RFNBYO57BSWI4GR5DAXAL34FYV.json","graph_json":"https://pith.science/api/pith-number/RFNBYO57BSWI4GR5DAXAL34FYV/graph.json","events_json":"https://pith.science/api/pith-number/RFNBYO57BSWI4GR5DAXAL34FYV/events.json","paper":"https://pith.science/paper/RFNBYO57"},"agent_actions":{"view_html":"https://pith.science/pith/RFNBYO57BSWI4GR5DAXAL34FYV","download_json":"https://pith.science/pith/RFNBYO57BSWI4GR5DAXAL34FYV.json","view_paper":"https://pith.science/paper/RFNBYO57","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2302.03655&json=true","fetch_graph":"https://pith.science/api/pith-number/RFNBYO57BSWI4GR5DAXAL34FYV/graph.json","fetch_events":"https://pith.science/api/pith-number/RFNBYO57BSWI4GR5DAXAL34FYV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RFNBYO57BSWI4GR5DAXAL34FYV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RFNBYO57BSWI4GR5DAXAL34FYV/action/storage_attestation","attest_author":"https://pith.science/pith/RFNBYO57BSWI4GR5DAXAL34FYV/action/author_attestation","sign_citation":"https://pith.science/pith/RFNBYO57BSWI4GR5DAXAL34FYV/action/citation_signature","submit_replication":"https://pith.science/pith/RFNBYO57BSWI4GR5DAXAL34FYV/action/replication_record"}},"created_at":"2026-07-05T06:20:40.562669+00:00","updated_at":"2026-07-05T06:20:40.562669+00:00"}