{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:452RKRRE54DD4T2OMDH65DDQX6","short_pith_number":"pith:452RKRRE","schema_version":"1.0","canonical_sha256":"e775154624ef063e4f4e60cfee8c70bf84042133b92733370703392b91781014","source":{"kind":"arxiv","id":"2209.15323","version":2},"attestation_state":"computed","paper":{"title":"SmallCap: Lightweight Image Captioning Prompted with Retrieval Augmentation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Bruno Martins, Desmond Elliott, Rita Ramos, Yova Kementchedjhieva","submitted_at":"2022-09-30T09:03:22Z","abstract_excerpt":"Recent advances in image captioning have focused on scaling the data and model size, substantially increasing the cost of pre-training and finetuning. As an alternative to large models, we present SmallCap, which generates a caption conditioned on an input image and related captions retrieved from a datastore. Our model is lightweight and fast to train, as the only learned parameters are in newly introduced cross-attention layers between a pre-trained CLIP encoder and GPT-2 decoder. SmallCap can transfer to new domains without additional finetuning and can exploit large-scale data in a trainin"},"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":"2209.15323","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2022-09-30T09:03:22Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"7a4a0712f599adc7bb3d68a170e7277211c67196d8f31dd592e707c35fcfdb20","abstract_canon_sha256":"4a937ac5d6777c16dc3f74426a27a95bd2e2c3c91d3705838a88a266cb035c13"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:55:59.526331Z","signature_b64":"wWIpggO4QcAvV751ecuH57ynfl2rDJ40Yqx5fJ6iGkY2r2tmDjvdyU2bq/nLN9LEkxWKNZAIYYU3oO2KvYhIDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e775154624ef063e4f4e60cfee8c70bf84042133b92733370703392b91781014","last_reissued_at":"2026-07-05T05:55:59.525814Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:55:59.525814Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SmallCap: Lightweight Image Captioning Prompted with Retrieval Augmentation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Bruno Martins, Desmond Elliott, Rita Ramos, Yova Kementchedjhieva","submitted_at":"2022-09-30T09:03:22Z","abstract_excerpt":"Recent advances in image captioning have focused on scaling the data and model size, substantially increasing the cost of pre-training and finetuning. As an alternative to large models, we present SmallCap, which generates a caption conditioned on an input image and related captions retrieved from a datastore. Our model is lightweight and fast to train, as the only learned parameters are in newly introduced cross-attention layers between a pre-trained CLIP encoder and GPT-2 decoder. SmallCap can transfer to new domains without additional finetuning and can exploit large-scale data in a trainin"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2209.15323","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/2209.15323/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":"2209.15323","created_at":"2026-07-05T05:55:59.525870+00:00"},{"alias_kind":"arxiv_version","alias_value":"2209.15323v2","created_at":"2026-07-05T05:55:59.525870+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2209.15323","created_at":"2026-07-05T05:55:59.525870+00:00"},{"alias_kind":"pith_short_12","alias_value":"452RKRRE54DD","created_at":"2026-07-05T05:55:59.525870+00:00"},{"alias_kind":"pith_short_16","alias_value":"452RKRRE54DD4T2O","created_at":"2026-07-05T05:55:59.525870+00:00"},{"alias_kind":"pith_short_8","alias_value":"452RKRRE","created_at":"2026-07-05T05:55:59.525870+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2412.19289","citing_title":"ViPCap: Retrieval Text-Based Visual Prompts for Lightweight Image Captioning","ref_index":38,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/452RKRRE54DD4T2OMDH65DDQX6","json":"https://pith.science/pith/452RKRRE54DD4T2OMDH65DDQX6.json","graph_json":"https://pith.science/api/pith-number/452RKRRE54DD4T2OMDH65DDQX6/graph.json","events_json":"https://pith.science/api/pith-number/452RKRRE54DD4T2OMDH65DDQX6/events.json","paper":"https://pith.science/paper/452RKRRE"},"agent_actions":{"view_html":"https://pith.science/pith/452RKRRE54DD4T2OMDH65DDQX6","download_json":"https://pith.science/pith/452RKRRE54DD4T2OMDH65DDQX6.json","view_paper":"https://pith.science/paper/452RKRRE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2209.15323&json=true","fetch_graph":"https://pith.science/api/pith-number/452RKRRE54DD4T2OMDH65DDQX6/graph.json","fetch_events":"https://pith.science/api/pith-number/452RKRRE54DD4T2OMDH65DDQX6/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/452RKRRE54DD4T2OMDH65DDQX6/action/timestamp_anchor","attest_storage":"https://pith.science/pith/452RKRRE54DD4T2OMDH65DDQX6/action/storage_attestation","attest_author":"https://pith.science/pith/452RKRRE54DD4T2OMDH65DDQX6/action/author_attestation","sign_citation":"https://pith.science/pith/452RKRRE54DD4T2OMDH65DDQX6/action/citation_signature","submit_replication":"https://pith.science/pith/452RKRRE54DD4T2OMDH65DDQX6/action/replication_record"}},"created_at":"2026-07-05T05:55:59.525870+00:00","updated_at":"2026-07-05T05:55:59.525870+00:00"}