{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:IVC6L2EPSEZ3XMBELJL6UZZ572","short_pith_number":"pith:IVC6L2EP","schema_version":"1.0","canonical_sha256":"4545e5e88f9133bbb0245a57ea673dfe8c17a55c4043af7691088bfb683f4085","source":{"kind":"arxiv","id":"2303.17561","version":2},"attestation_state":"computed","paper":{"title":"SoftCLIP: Softer Cross-modal Alignment Makes CLIP Stronger","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Jie Yang, Jinfeng Liu, Ke Li, Tong Wu Enwei Zhang, Wei Liu, Xing Sun, Yuting Gao, Zihan Xu","submitted_at":"2023-03-30T17:27:22Z","abstract_excerpt":"During the preceding biennium, vision-language pre-training has achieved noteworthy success on several downstream tasks. Nevertheless, acquiring high-quality image-text pairs, where the pairs are entirely exclusive of each other, remains a challenging task, and noise exists in the commonly used datasets. To address this issue, we propose SoftCLIP, a novel approach that relaxes the strict one-to-one constraint and achieves a soft cross-modal alignment by introducing a softened target, which is generated from the fine-grained intra-modal self-similarity. The intra-modal guidance is indicative to"},"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":"2303.17561","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-03-30T17:27:22Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"cc019d4582f9baa9fa3ffdde1923c03467f0727798625318419beab1b135e16c","abstract_canon_sha256":"67663fc70da771fc08d705993a9070b6525a6a3e2e74e421705c095c79b0ed3e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:24:49.130637Z","signature_b64":"dJNBo4/uC1uAT4IR+n6OErfOv8TuAWj0p2HTyqzWyy9qvQ92ay5hATZLyJPzckWnpM9Qsk2sKjBVrigCzTAICQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4545e5e88f9133bbb0245a57ea673dfe8c17a55c4043af7691088bfb683f4085","last_reissued_at":"2026-07-05T07:24:49.130189Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:24:49.130189Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SoftCLIP: Softer Cross-modal Alignment Makes CLIP Stronger","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Jie Yang, Jinfeng Liu, Ke Li, Tong Wu Enwei Zhang, Wei Liu, Xing Sun, Yuting Gao, Zihan Xu","submitted_at":"2023-03-30T17:27:22Z","abstract_excerpt":"During the preceding biennium, vision-language pre-training has achieved noteworthy success on several downstream tasks. Nevertheless, acquiring high-quality image-text pairs, where the pairs are entirely exclusive of each other, remains a challenging task, and noise exists in the commonly used datasets. To address this issue, we propose SoftCLIP, a novel approach that relaxes the strict one-to-one constraint and achieves a soft cross-modal alignment by introducing a softened target, which is generated from the fine-grained intra-modal self-similarity. The intra-modal guidance is indicative to"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2303.17561","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/2303.17561/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":"2303.17561","created_at":"2026-07-05T07:24:49.130244+00:00"},{"alias_kind":"arxiv_version","alias_value":"2303.17561v2","created_at":"2026-07-05T07:24:49.130244+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2303.17561","created_at":"2026-07-05T07:24:49.130244+00:00"},{"alias_kind":"pith_short_12","alias_value":"IVC6L2EPSEZ3","created_at":"2026-07-05T07:24:49.130244+00:00"},{"alias_kind":"pith_short_16","alias_value":"IVC6L2EPSEZ3XMBE","created_at":"2026-07-05T07:24:49.130244+00:00"},{"alias_kind":"pith_short_8","alias_value":"IVC6L2EP","created_at":"2026-07-05T07:24:49.130244+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2502.08254","citing_title":"UniCoRN: Unified Commented Retrieval Network with LMMs","ref_index":26,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IVC6L2EPSEZ3XMBELJL6UZZ572","json":"https://pith.science/pith/IVC6L2EPSEZ3XMBELJL6UZZ572.json","graph_json":"https://pith.science/api/pith-number/IVC6L2EPSEZ3XMBELJL6UZZ572/graph.json","events_json":"https://pith.science/api/pith-number/IVC6L2EPSEZ3XMBELJL6UZZ572/events.json","paper":"https://pith.science/paper/IVC6L2EP"},"agent_actions":{"view_html":"https://pith.science/pith/IVC6L2EPSEZ3XMBELJL6UZZ572","download_json":"https://pith.science/pith/IVC6L2EPSEZ3XMBELJL6UZZ572.json","view_paper":"https://pith.science/paper/IVC6L2EP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2303.17561&json=true","fetch_graph":"https://pith.science/api/pith-number/IVC6L2EPSEZ3XMBELJL6UZZ572/graph.json","fetch_events":"https://pith.science/api/pith-number/IVC6L2EPSEZ3XMBELJL6UZZ572/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IVC6L2EPSEZ3XMBELJL6UZZ572/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IVC6L2EPSEZ3XMBELJL6UZZ572/action/storage_attestation","attest_author":"https://pith.science/pith/IVC6L2EPSEZ3XMBELJL6UZZ572/action/author_attestation","sign_citation":"https://pith.science/pith/IVC6L2EPSEZ3XMBELJL6UZZ572/action/citation_signature","submit_replication":"https://pith.science/pith/IVC6L2EPSEZ3XMBELJL6UZZ572/action/replication_record"}},"created_at":"2026-07-05T07:24:49.130244+00:00","updated_at":"2026-07-05T07:24:49.130244+00:00"}