{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:DUZ5DK7746B34UCI2UQIHWHGZJ","short_pith_number":"pith:DUZ5DK77","schema_version":"1.0","canonical_sha256":"1d33d1abffe783be5048d52083d8e6ca736c5642926b3667f8b4d8de7bdf7f06","source":{"kind":"arxiv","id":"2502.02202","version":2},"attestation_state":"computed","paper":{"title":"Multi-level Supervised Contrastive Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Barbod Pajoum, Harshit Rawal, Marius Kloft, Naghmeh Ghanooni, Sophie Fellenz, Vo Nguyen Le Duy","submitted_at":"2025-02-04T10:27:54Z","abstract_excerpt":"Contrastive learning is a well-established paradigm in representation learning. The standard framework of contrastive learning minimizes the distance between \"similar\" instances and maximizes the distance between dissimilar ones in the projection space, disregarding the various aspects of similarity that can exist between two samples. Current methods rely on a single projection head, which fails to capture the full complexity of different aspects of a sample, leading to suboptimal performance, especially in scenarios with limited training data. In this paper, we present a novel supervised cont"},"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":"2502.02202","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-04T10:27:54Z","cross_cats_sorted":[],"title_canon_sha256":"6a7105ff016599497e7a1a55dbfad8abb1710a62578e647f0c4c31a7c90b63dc","abstract_canon_sha256":"a30468c08b1a5a9d36a17c387c8c83703697a9b74d53626818dfdf7e99d61b1c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:09:52.896571Z","signature_b64":"DROIp+ypLv/JNHh5bXW0R18XcfgUFiKts9dky4relZq6UCtn0sXxiQKzz3PozftEitoNoSEDCPSbEIJzFUFQAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1d33d1abffe783be5048d52083d8e6ca736c5642926b3667f8b4d8de7bdf7f06","last_reissued_at":"2026-07-05T10:09:52.896157Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:09:52.896157Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Multi-level Supervised Contrastive Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Barbod Pajoum, Harshit Rawal, Marius Kloft, Naghmeh Ghanooni, Sophie Fellenz, Vo Nguyen Le Duy","submitted_at":"2025-02-04T10:27:54Z","abstract_excerpt":"Contrastive learning is a well-established paradigm in representation learning. The standard framework of contrastive learning minimizes the distance between \"similar\" instances and maximizes the distance between dissimilar ones in the projection space, disregarding the various aspects of similarity that can exist between two samples. Current methods rely on a single projection head, which fails to capture the full complexity of different aspects of a sample, leading to suboptimal performance, especially in scenarios with limited training data. In this paper, we present a novel supervised cont"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.02202","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/2502.02202/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":"2502.02202","created_at":"2026-07-05T10:09:52.896208+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.02202v2","created_at":"2026-07-05T10:09:52.896208+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.02202","created_at":"2026-07-05T10:09:52.896208+00:00"},{"alias_kind":"pith_short_12","alias_value":"DUZ5DK7746B3","created_at":"2026-07-05T10:09:52.896208+00:00"},{"alias_kind":"pith_short_16","alias_value":"DUZ5DK7746B34UCI","created_at":"2026-07-05T10:09:52.896208+00:00"},{"alias_kind":"pith_short_8","alias_value":"DUZ5DK77","created_at":"2026-07-05T10:09:52.896208+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.21838","citing_title":"Beyond Flat Labels: Level-Restricted Contrastive Learning for Hierarchical Fine-Grained Vision Classification","ref_index":7,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DUZ5DK7746B34UCI2UQIHWHGZJ","json":"https://pith.science/pith/DUZ5DK7746B34UCI2UQIHWHGZJ.json","graph_json":"https://pith.science/api/pith-number/DUZ5DK7746B34UCI2UQIHWHGZJ/graph.json","events_json":"https://pith.science/api/pith-number/DUZ5DK7746B34UCI2UQIHWHGZJ/events.json","paper":"https://pith.science/paper/DUZ5DK77"},"agent_actions":{"view_html":"https://pith.science/pith/DUZ5DK7746B34UCI2UQIHWHGZJ","download_json":"https://pith.science/pith/DUZ5DK7746B34UCI2UQIHWHGZJ.json","view_paper":"https://pith.science/paper/DUZ5DK77","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.02202&json=true","fetch_graph":"https://pith.science/api/pith-number/DUZ5DK7746B34UCI2UQIHWHGZJ/graph.json","fetch_events":"https://pith.science/api/pith-number/DUZ5DK7746B34UCI2UQIHWHGZJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DUZ5DK7746B34UCI2UQIHWHGZJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DUZ5DK7746B34UCI2UQIHWHGZJ/action/storage_attestation","attest_author":"https://pith.science/pith/DUZ5DK7746B34UCI2UQIHWHGZJ/action/author_attestation","sign_citation":"https://pith.science/pith/DUZ5DK7746B34UCI2UQIHWHGZJ/action/citation_signature","submit_replication":"https://pith.science/pith/DUZ5DK7746B34UCI2UQIHWHGZJ/action/replication_record"}},"created_at":"2026-07-05T10:09:52.896208+00:00","updated_at":"2026-07-05T10:09:52.896208+00:00"}