{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:B5MXNKQS7XSKKB4VXAVK6UA3WA","short_pith_number":"pith:B5MXNKQS","schema_version":"1.0","canonical_sha256":"0f5976aa12fde4a50795b82aaf501bb02f78ef4c664de1c1e9aced11edc375f0","source":{"kind":"arxiv","id":"2502.19665","version":2},"attestation_state":"computed","paper":{"title":"Out-of-distribution Generalization for Total Variation based Invariant Risk Minimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Tianqi Zhong, Yuanchao Wang, Zhao-Rong Lai","submitted_at":"2025-02-27T01:11:11Z","abstract_excerpt":"Invariant risk minimization is an important general machine learning framework that has recently been interpreted as a total variation model (IRM-TV). However, how to improve out-of-distribution (OOD) generalization in the IRM-TV setting remains unsolved. In this paper, we extend IRM-TV to a Lagrangian multiplier model named OOD-TV-IRM. We find that the autonomous TV penalty hyperparameter is exactly the Lagrangian multiplier. Thus OOD-TV-IRM is essentially a primal-dual optimization model, where the primal optimization minimizes the entire invariant risk and the dual optimization strengthens "},"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":"2502.19665","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-27T01:11:11Z","cross_cats_sorted":[],"title_canon_sha256":"c25edf9379e094193750962bcc4814f35e3887ad5874bbf9aba79cff51815785","abstract_canon_sha256":"f1df17160c915fe3bf807c57935a15781bee1528f09e496ae311778409a96687"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:21:31.767878Z","signature_b64":"fgw4ZtOQrsDPyJ6oEYPgViNqTWwk839B0zuYpyGLPFMEJZYyOUTZNt4XsR3+PikSLMCExsZ8pyn5AgA2ScGrAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0f5976aa12fde4a50795b82aaf501bb02f78ef4c664de1c1e9aced11edc375f0","last_reissued_at":"2026-07-05T10:21:31.767291Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:21:31.767291Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Out-of-distribution Generalization for Total Variation based Invariant Risk Minimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Tianqi Zhong, Yuanchao Wang, Zhao-Rong Lai","submitted_at":"2025-02-27T01:11:11Z","abstract_excerpt":"Invariant risk minimization is an important general machine learning framework that has recently been interpreted as a total variation model (IRM-TV). However, how to improve out-of-distribution (OOD) generalization in the IRM-TV setting remains unsolved. In this paper, we extend IRM-TV to a Lagrangian multiplier model named OOD-TV-IRM. We find that the autonomous TV penalty hyperparameter is exactly the Lagrangian multiplier. Thus OOD-TV-IRM is essentially a primal-dual optimization model, where the primal optimization minimizes the entire invariant risk and the dual optimization strengthens "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.19665","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/2502.19665/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":"2502.19665","created_at":"2026-07-05T10:21:31.767366+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.19665v2","created_at":"2026-07-05T10:21:31.767366+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.19665","created_at":"2026-07-05T10:21:31.767366+00:00"},{"alias_kind":"pith_short_12","alias_value":"B5MXNKQS7XSK","created_at":"2026-07-05T10:21:31.767366+00:00"},{"alias_kind":"pith_short_16","alias_value":"B5MXNKQS7XSKKB4V","created_at":"2026-07-05T10:21:31.767366+00:00"},{"alias_kind":"pith_short_8","alias_value":"B5MXNKQS","created_at":"2026-07-05T10:21:31.767366+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.05335","citing_title":"Cross-Machine Anomaly Detection Leveraging Pre-trained Time-series Model","ref_index":24,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/B5MXNKQS7XSKKB4VXAVK6UA3WA","json":"https://pith.science/pith/B5MXNKQS7XSKKB4VXAVK6UA3WA.json","graph_json":"https://pith.science/api/pith-number/B5MXNKQS7XSKKB4VXAVK6UA3WA/graph.json","events_json":"https://pith.science/api/pith-number/B5MXNKQS7XSKKB4VXAVK6UA3WA/events.json","paper":"https://pith.science/paper/B5MXNKQS"},"agent_actions":{"view_html":"https://pith.science/pith/B5MXNKQS7XSKKB4VXAVK6UA3WA","download_json":"https://pith.science/pith/B5MXNKQS7XSKKB4VXAVK6UA3WA.json","view_paper":"https://pith.science/paper/B5MXNKQS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.19665&json=true","fetch_graph":"https://pith.science/api/pith-number/B5MXNKQS7XSKKB4VXAVK6UA3WA/graph.json","fetch_events":"https://pith.science/api/pith-number/B5MXNKQS7XSKKB4VXAVK6UA3WA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/B5MXNKQS7XSKKB4VXAVK6UA3WA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/B5MXNKQS7XSKKB4VXAVK6UA3WA/action/storage_attestation","attest_author":"https://pith.science/pith/B5MXNKQS7XSKKB4VXAVK6UA3WA/action/author_attestation","sign_citation":"https://pith.science/pith/B5MXNKQS7XSKKB4VXAVK6UA3WA/action/citation_signature","submit_replication":"https://pith.science/pith/B5MXNKQS7XSKKB4VXAVK6UA3WA/action/replication_record"}},"created_at":"2026-07-05T10:21:31.767366+00:00","updated_at":"2026-07-05T10:21:31.767366+00:00"}