{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:CVS43ANIJIHBR2TJ4QWQUSUT53","short_pith_number":"pith:CVS43ANI","schema_version":"1.0","canonical_sha256":"1565cd81a84a0e18ea69e42d0a4a93eecbb5b553397209b0dec66aaa4d5b5e23","source":{"kind":"arxiv","id":"2205.14459","version":2},"attestation_state":"computed","paper":{"title":"CyCLIP: Cyclic Contrastive Language-Image Pretraining","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Aditya Grover, Hritik Bansal, Ryan A. Rossi, Shashank Goel, Sumit Bhatia, Vishwa Vinay","submitted_at":"2022-05-28T15:31:17Z","abstract_excerpt":"Recent advances in contrastive representation learning over paired image-text data have led to models such as CLIP that achieve state-of-the-art performance for zero-shot classification and distributional robustness. Such models typically require joint reasoning in the image and text representation spaces for downstream inference tasks. Contrary to prior beliefs, we demonstrate that the image and text representations learned via a standard contrastive objective are not interchangeable and can lead to inconsistent downstream predictions. To mitigate this issue, we formalize consistency and prop"},"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":"2205.14459","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2022-05-28T15:31:17Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"9b318346cc2954e0db54f8cd499128348ecb8fd08db4562aea9694a7dbb6e372","abstract_canon_sha256":"573fc2fc5b0fd932a99ebe5262b038063a1b1f90f3f83fc7b88e58ff56302de5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:10:56.081749Z","signature_b64":"QHbbtahB3zZEVSdFCiWM9XnEc1FtESTjvuNvI68mjeIbRObupB6eaSUVryBKxFWAl9chgirRSql4pIV+WrOeAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1565cd81a84a0e18ea69e42d0a4a93eecbb5b553397209b0dec66aaa4d5b5e23","last_reissued_at":"2026-07-05T05:10:56.081339Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:10:56.081339Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CyCLIP: Cyclic Contrastive Language-Image Pretraining","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Aditya Grover, Hritik Bansal, Ryan A. Rossi, Shashank Goel, Sumit Bhatia, Vishwa Vinay","submitted_at":"2022-05-28T15:31:17Z","abstract_excerpt":"Recent advances in contrastive representation learning over paired image-text data have led to models such as CLIP that achieve state-of-the-art performance for zero-shot classification and distributional robustness. Such models typically require joint reasoning in the image and text representation spaces for downstream inference tasks. Contrary to prior beliefs, we demonstrate that the image and text representations learned via a standard contrastive objective are not interchangeable and can lead to inconsistent downstream predictions. To mitigate this issue, we formalize consistency and prop"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2205.14459","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/2205.14459/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":"2205.14459","created_at":"2026-07-05T05:10:56.081415+00:00"},{"alias_kind":"arxiv_version","alias_value":"2205.14459v2","created_at":"2026-07-05T05:10:56.081415+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2205.14459","created_at":"2026-07-05T05:10:56.081415+00:00"},{"alias_kind":"pith_short_12","alias_value":"CVS43ANIJIHB","created_at":"2026-07-05T05:10:56.081415+00:00"},{"alias_kind":"pith_short_16","alias_value":"CVS43ANIJIHBR2TJ","created_at":"2026-07-05T05:10:56.081415+00:00"},{"alias_kind":"pith_short_8","alias_value":"CVS43ANI","created_at":"2026-07-05T05:10:56.081415+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CVS43ANIJIHBR2TJ4QWQUSUT53","json":"https://pith.science/pith/CVS43ANIJIHBR2TJ4QWQUSUT53.json","graph_json":"https://pith.science/api/pith-number/CVS43ANIJIHBR2TJ4QWQUSUT53/graph.json","events_json":"https://pith.science/api/pith-number/CVS43ANIJIHBR2TJ4QWQUSUT53/events.json","paper":"https://pith.science/paper/CVS43ANI"},"agent_actions":{"view_html":"https://pith.science/pith/CVS43ANIJIHBR2TJ4QWQUSUT53","download_json":"https://pith.science/pith/CVS43ANIJIHBR2TJ4QWQUSUT53.json","view_paper":"https://pith.science/paper/CVS43ANI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2205.14459&json=true","fetch_graph":"https://pith.science/api/pith-number/CVS43ANIJIHBR2TJ4QWQUSUT53/graph.json","fetch_events":"https://pith.science/api/pith-number/CVS43ANIJIHBR2TJ4QWQUSUT53/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CVS43ANIJIHBR2TJ4QWQUSUT53/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CVS43ANIJIHBR2TJ4QWQUSUT53/action/storage_attestation","attest_author":"https://pith.science/pith/CVS43ANIJIHBR2TJ4QWQUSUT53/action/author_attestation","sign_citation":"https://pith.science/pith/CVS43ANIJIHBR2TJ4QWQUSUT53/action/citation_signature","submit_replication":"https://pith.science/pith/CVS43ANIJIHBR2TJ4QWQUSUT53/action/replication_record"}},"created_at":"2026-07-05T05:10:56.081415+00:00","updated_at":"2026-07-05T05:10:56.081415+00:00"}