{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:QPYHCV7PXVOREMEAKO3SRVXVFH","short_pith_number":"pith:QPYHCV7P","schema_version":"1.0","canonical_sha256":"83f07157efbd5d12308053b728d6f529cbdd634faaa9cc8e08920f394faabbfb","source":{"kind":"arxiv","id":"2404.09387","version":3},"attestation_state":"computed","paper":{"title":"RankCLIP: Ranking-Consistent Language-Image Pretraining","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Yiming Zhang, Yining Sun, Zenghui Ding, Zhaorun Chen, Zhili Feng, Zhuokai Zhao","submitted_at":"2024-04-15T00:12:27Z","abstract_excerpt":"Self-supervised contrastive learning models, such as CLIP, have set new benchmarks for vision-language models in many downstream tasks. However, their dependency on rigid one-to-one mappings overlooks the complex and often multifaceted relationships between and within texts and images. To this end, we introduce RankCLIP, a novel pre-training method that extends beyond the rigid one-to-one matching framework of CLIP and its variants. By extending the traditional pair-wise loss to list-wise, and leveraging both in-modal and cross-modal ranking consistency, RankCLIP improves the alignment process"},"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":"2404.09387","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-04-15T00:12:27Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"bae7625089838d25fc24b4de391da4c9242e2f0dbf6307a15ea1ec974149ad4d","abstract_canon_sha256":"25cc7e6aa28bccc721708e3c48641dd0af6b03d9292af3e01e1bae09d38f679f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:38:20.640639Z","signature_b64":"gCcgCce5QTS03fWTVmbkCfOhBpVb7Zfg8W/jtX404vhrk0VLEV7w1LTTJroiMIHFXZikHVgKTeroF/mDqvWHBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"83f07157efbd5d12308053b728d6f529cbdd634faaa9cc8e08920f394faabbfb","last_reissued_at":"2026-07-05T10:38:20.640156Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:38:20.640156Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"RankCLIP: Ranking-Consistent Language-Image Pretraining","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Yiming Zhang, Yining Sun, Zenghui Ding, Zhaorun Chen, Zhili Feng, Zhuokai Zhao","submitted_at":"2024-04-15T00:12:27Z","abstract_excerpt":"Self-supervised contrastive learning models, such as CLIP, have set new benchmarks for vision-language models in many downstream tasks. However, their dependency on rigid one-to-one mappings overlooks the complex and often multifaceted relationships between and within texts and images. To this end, we introduce RankCLIP, a novel pre-training method that extends beyond the rigid one-to-one matching framework of CLIP and its variants. By extending the traditional pair-wise loss to list-wise, and leveraging both in-modal and cross-modal ranking consistency, RankCLIP improves the alignment process"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.09387","kind":"arxiv","version":3},"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/2404.09387/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":"2404.09387","created_at":"2026-07-05T10:38:20.640211+00:00"},{"alias_kind":"arxiv_version","alias_value":"2404.09387v3","created_at":"2026-07-05T10:38:20.640211+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.09387","created_at":"2026-07-05T10:38:20.640211+00:00"},{"alias_kind":"pith_short_12","alias_value":"QPYHCV7PXVOR","created_at":"2026-07-05T10:38:20.640211+00:00"},{"alias_kind":"pith_short_16","alias_value":"QPYHCV7PXVOREMEA","created_at":"2026-07-05T10:38:20.640211+00:00"},{"alias_kind":"pith_short_8","alias_value":"QPYHCV7P","created_at":"2026-07-05T10:38:20.640211+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.06592","citing_title":"DINORANKCLIP: DINOv3 Distillation and Injection for Vision-Language Pretraining with High-Order Ranking Consistency","ref_index":19,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QPYHCV7PXVOREMEAKO3SRVXVFH","json":"https://pith.science/pith/QPYHCV7PXVOREMEAKO3SRVXVFH.json","graph_json":"https://pith.science/api/pith-number/QPYHCV7PXVOREMEAKO3SRVXVFH/graph.json","events_json":"https://pith.science/api/pith-number/QPYHCV7PXVOREMEAKO3SRVXVFH/events.json","paper":"https://pith.science/paper/QPYHCV7P"},"agent_actions":{"view_html":"https://pith.science/pith/QPYHCV7PXVOREMEAKO3SRVXVFH","download_json":"https://pith.science/pith/QPYHCV7PXVOREMEAKO3SRVXVFH.json","view_paper":"https://pith.science/paper/QPYHCV7P","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2404.09387&json=true","fetch_graph":"https://pith.science/api/pith-number/QPYHCV7PXVOREMEAKO3SRVXVFH/graph.json","fetch_events":"https://pith.science/api/pith-number/QPYHCV7PXVOREMEAKO3SRVXVFH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QPYHCV7PXVOREMEAKO3SRVXVFH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QPYHCV7PXVOREMEAKO3SRVXVFH/action/storage_attestation","attest_author":"https://pith.science/pith/QPYHCV7PXVOREMEAKO3SRVXVFH/action/author_attestation","sign_citation":"https://pith.science/pith/QPYHCV7PXVOREMEAKO3SRVXVFH/action/citation_signature","submit_replication":"https://pith.science/pith/QPYHCV7PXVOREMEAKO3SRVXVFH/action/replication_record"}},"created_at":"2026-07-05T10:38:20.640211+00:00","updated_at":"2026-07-05T10:38:20.640211+00:00"}