{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:XHAVVRZW52IAQUACEDIKJNCZDM","short_pith_number":"pith:XHAVVRZW","schema_version":"1.0","canonical_sha256":"b9c15ac736ee9008500220d0a4b4591b3032d74566cccdd24666588a1a5b18c7","source":{"kind":"arxiv","id":"2506.13714","version":1},"attestation_state":"computed","paper":{"title":"Understanding Learning Invariance in Deep Linear Networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","math.ST","stat.TH"],"primary_cat":"stat.ML","authors_text":"Guido Mont\\'ufar, Hao Duan","submitted_at":"2025-06-16T17:24:07Z","abstract_excerpt":"Equivariant and invariant machine learning models exploit symmetries and structural patterns in data to improve sample efficiency. While empirical studies suggest that data-driven methods such as regularization and data augmentation can perform comparably to explicitly invariant models, theoretical insights remain scarce. In this paper, we provide a theoretical comparison of three approaches for achieving invariance: data augmentation, regularization, and hard-wiring. We focus on mean squared error regression with deep linear networks, which parametrize rank-bounded linear maps and can be hard"},"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":"2506.13714","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2025-06-16T17:24:07Z","cross_cats_sorted":["cs.LG","math.ST","stat.TH"],"title_canon_sha256":"af585bda2ba4a6dbea0feb729a0e16ba4885e342a6465976acfbe88827c57b61","abstract_canon_sha256":"82d77191cda0a7f97fe2c1048f9cc74f94649f4a6a27f3674e1a7b3c513a3944"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:22:25.422030Z","signature_b64":"DcvqVHL4roT5oRHwbePK+BEFOMhTGwdDcMOuxHwtugrYUAIE3WEYNd3w00yHK4DxScfZ4izajff8gv5OTqFJCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b9c15ac736ee9008500220d0a4b4591b3032d74566cccdd24666588a1a5b18c7","last_reissued_at":"2026-07-05T11:22:25.421502Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:22:25.421502Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Understanding Learning Invariance in Deep Linear Networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","math.ST","stat.TH"],"primary_cat":"stat.ML","authors_text":"Guido Mont\\'ufar, Hao Duan","submitted_at":"2025-06-16T17:24:07Z","abstract_excerpt":"Equivariant and invariant machine learning models exploit symmetries and structural patterns in data to improve sample efficiency. While empirical studies suggest that data-driven methods such as regularization and data augmentation can perform comparably to explicitly invariant models, theoretical insights remain scarce. In this paper, we provide a theoretical comparison of three approaches for achieving invariance: data augmentation, regularization, and hard-wiring. We focus on mean squared error regression with deep linear networks, which parametrize rank-bounded linear maps and can be hard"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.13714","kind":"arxiv","version":1},"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/2506.13714/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":"2506.13714","created_at":"2026-07-05T11:22:25.421569+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.13714v1","created_at":"2026-07-05T11:22:25.421569+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.13714","created_at":"2026-07-05T11:22:25.421569+00:00"},{"alias_kind":"pith_short_12","alias_value":"XHAVVRZW52IA","created_at":"2026-07-05T11:22:25.421569+00:00"},{"alias_kind":"pith_short_16","alias_value":"XHAVVRZW52IAQUAC","created_at":"2026-07-05T11:22:25.421569+00:00"},{"alias_kind":"pith_short_8","alias_value":"XHAVVRZW","created_at":"2026-07-05T11:22:25.421569+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26273","citing_title":"Equivariance and Augmentation for Bayesian Neural Networks","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2606.10913","citing_title":"Conservation Laws from Data Symmetry in Neural Networks","ref_index":19,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XHAVVRZW52IAQUACEDIKJNCZDM","json":"https://pith.science/pith/XHAVVRZW52IAQUACEDIKJNCZDM.json","graph_json":"https://pith.science/api/pith-number/XHAVVRZW52IAQUACEDIKJNCZDM/graph.json","events_json":"https://pith.science/api/pith-number/XHAVVRZW52IAQUACEDIKJNCZDM/events.json","paper":"https://pith.science/paper/XHAVVRZW"},"agent_actions":{"view_html":"https://pith.science/pith/XHAVVRZW52IAQUACEDIKJNCZDM","download_json":"https://pith.science/pith/XHAVVRZW52IAQUACEDIKJNCZDM.json","view_paper":"https://pith.science/paper/XHAVVRZW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.13714&json=true","fetch_graph":"https://pith.science/api/pith-number/XHAVVRZW52IAQUACEDIKJNCZDM/graph.json","fetch_events":"https://pith.science/api/pith-number/XHAVVRZW52IAQUACEDIKJNCZDM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XHAVVRZW52IAQUACEDIKJNCZDM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XHAVVRZW52IAQUACEDIKJNCZDM/action/storage_attestation","attest_author":"https://pith.science/pith/XHAVVRZW52IAQUACEDIKJNCZDM/action/author_attestation","sign_citation":"https://pith.science/pith/XHAVVRZW52IAQUACEDIKJNCZDM/action/citation_signature","submit_replication":"https://pith.science/pith/XHAVVRZW52IAQUACEDIKJNCZDM/action/replication_record"}},"created_at":"2026-07-05T11:22:25.421569+00:00","updated_at":"2026-07-05T11:22:25.421569+00:00"}