{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:PEWUDBMX5VMXA5CZQHQZAUI3DG","short_pith_number":"pith:PEWUDBMX","schema_version":"1.0","canonical_sha256":"792d418597ed5970745981e190511b19a0e6e9f2e91e72c20128871063a6d8a1","source":{"kind":"arxiv","id":"2304.13273","version":3},"attestation_state":"computed","paper":{"title":"From Association to Generation: Text-only Captioning by Unsupervised Cross-modal Mapping","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.CV","authors_text":"Jitao Sang, Junyang Wang, Ming Yan, Yi Zhang","submitted_at":"2023-04-26T04:06:20Z","abstract_excerpt":"With the development of Vision-Language Pre-training Models (VLPMs) represented by CLIP and ALIGN, significant breakthroughs have been achieved for association-based visual tasks such as image classification and image-text retrieval by the zero-shot capability of CLIP without fine-tuning. However, CLIP is hard to apply to generation-based tasks. This is due to the lack of decoder architecture and pre-training tasks for generation. Although previous works have created generation capacity for CLIP through additional language models, a modality gap between the CLIP representations of different mo"},"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":"2304.13273","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-04-26T04:06:20Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"d49c3aba65521fbb480458413bc00fc15ca7911757809cad90703f7aba3821bb","abstract_canon_sha256":"645d3b76d49f26a173e2cda123a6e428b070703aa37198c9f351b1790585f995"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:07:38.942147Z","signature_b64":"f8rRDETLE1ZyaV+9CewbgPwsJyjkVNtckgiN8wVM4yZK4HsFLKUg51JXxJ7Ttb6tkMMAbxvWUlkhM+j9K6B1Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"792d418597ed5970745981e190511b19a0e6e9f2e91e72c20128871063a6d8a1","last_reissued_at":"2026-07-05T06:07:38.941752Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:07:38.941752Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"From Association to Generation: Text-only Captioning by Unsupervised Cross-modal Mapping","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.CV","authors_text":"Jitao Sang, Junyang Wang, Ming Yan, Yi Zhang","submitted_at":"2023-04-26T04:06:20Z","abstract_excerpt":"With the development of Vision-Language Pre-training Models (VLPMs) represented by CLIP and ALIGN, significant breakthroughs have been achieved for association-based visual tasks such as image classification and image-text retrieval by the zero-shot capability of CLIP without fine-tuning. However, CLIP is hard to apply to generation-based tasks. This is due to the lack of decoder architecture and pre-training tasks for generation. Although previous works have created generation capacity for CLIP through additional language models, a modality gap between the CLIP representations of different mo"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2304.13273","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/2304.13273/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":"2304.13273","created_at":"2026-07-05T06:07:38.941815+00:00"},{"alias_kind":"arxiv_version","alias_value":"2304.13273v3","created_at":"2026-07-05T06:07:38.941815+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2304.13273","created_at":"2026-07-05T06:07:38.941815+00:00"},{"alias_kind":"pith_short_12","alias_value":"PEWUDBMX5VMX","created_at":"2026-07-05T06:07:38.941815+00:00"},{"alias_kind":"pith_short_16","alias_value":"PEWUDBMX5VMXA5CZ","created_at":"2026-07-05T06:07:38.941815+00:00"},{"alias_kind":"pith_short_8","alias_value":"PEWUDBMX","created_at":"2026-07-05T06:07:38.941815+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2412.08158","citing_title":"How Vision-Language Tasks Benefit from Large Pre-trained Models: A Survey","ref_index":79,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PEWUDBMX5VMXA5CZQHQZAUI3DG","json":"https://pith.science/pith/PEWUDBMX5VMXA5CZQHQZAUI3DG.json","graph_json":"https://pith.science/api/pith-number/PEWUDBMX5VMXA5CZQHQZAUI3DG/graph.json","events_json":"https://pith.science/api/pith-number/PEWUDBMX5VMXA5CZQHQZAUI3DG/events.json","paper":"https://pith.science/paper/PEWUDBMX"},"agent_actions":{"view_html":"https://pith.science/pith/PEWUDBMX5VMXA5CZQHQZAUI3DG","download_json":"https://pith.science/pith/PEWUDBMX5VMXA5CZQHQZAUI3DG.json","view_paper":"https://pith.science/paper/PEWUDBMX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2304.13273&json=true","fetch_graph":"https://pith.science/api/pith-number/PEWUDBMX5VMXA5CZQHQZAUI3DG/graph.json","fetch_events":"https://pith.science/api/pith-number/PEWUDBMX5VMXA5CZQHQZAUI3DG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PEWUDBMX5VMXA5CZQHQZAUI3DG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PEWUDBMX5VMXA5CZQHQZAUI3DG/action/storage_attestation","attest_author":"https://pith.science/pith/PEWUDBMX5VMXA5CZQHQZAUI3DG/action/author_attestation","sign_citation":"https://pith.science/pith/PEWUDBMX5VMXA5CZQHQZAUI3DG/action/citation_signature","submit_replication":"https://pith.science/pith/PEWUDBMX5VMXA5CZQHQZAUI3DG/action/replication_record"}},"created_at":"2026-07-05T06:07:38.941815+00:00","updated_at":"2026-07-05T06:07:38.941815+00:00"}