{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:DHU4QZFBXADYUWJIOXJ6W4U2MF","short_pith_number":"pith:DHU4QZFB","schema_version":"1.0","canonical_sha256":"19e9c864a1b8078a592875d3eb729a61771c37239b946b9ae0cf0c097998b21b","source":{"kind":"arxiv","id":"2305.19821","version":1},"attestation_state":"computed","paper":{"title":"LMCap: Few-shot Multilingual Image Captioning by Retrieval Augmented Language Model Prompting","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.CL","authors_text":"Bruno Martins, Desmond Elliott, Rita Ramos","submitted_at":"2023-05-31T13:03:17Z","abstract_excerpt":"Multilingual image captioning has recently been tackled by training with large-scale machine translated data, which is an expensive, noisy, and time-consuming process. Without requiring any multilingual caption data, we propose LMCap, an image-blind few-shot multilingual captioning model that works by prompting a language model with retrieved captions. Specifically, instead of following the standard encoder-decoder paradigm, given an image, LMCap first retrieves the captions of similar images using a multilingual CLIP encoder. These captions are then combined into a prompt for an XGLM decoder,"},"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":"2305.19821","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-05-31T13:03:17Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"c0ab894622ee1f3a12bcfb2f651b127dc0e14490bf4894b3ab1108b94d6442f8","abstract_canon_sha256":"6f88627c69571b6c76b6c2d6e7e3d6ff280adf93ecf6cc9296e65f517590b7cd"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:16:03.649665Z","signature_b64":"nopLtykCrzYmI+kd/UsgtwcCD9HDm0mcRXZhxAjIxaVQwc1bmRN2Gf2EtEvS3yzNeidJS16nhgmLzJG/FOeFAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"19e9c864a1b8078a592875d3eb729a61771c37239b946b9ae0cf0c097998b21b","last_reissued_at":"2026-07-05T06:16:03.649091Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:16:03.649091Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LMCap: Few-shot Multilingual Image Captioning by Retrieval Augmented Language Model Prompting","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.CL","authors_text":"Bruno Martins, Desmond Elliott, Rita Ramos","submitted_at":"2023-05-31T13:03:17Z","abstract_excerpt":"Multilingual image captioning has recently been tackled by training with large-scale machine translated data, which is an expensive, noisy, and time-consuming process. Without requiring any multilingual caption data, we propose LMCap, an image-blind few-shot multilingual captioning model that works by prompting a language model with retrieved captions. Specifically, instead of following the standard encoder-decoder paradigm, given an image, LMCap first retrieves the captions of similar images using a multilingual CLIP encoder. These captions are then combined into a prompt for an XGLM decoder,"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.19821","kind":"arxiv","version":1},"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/2305.19821/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":"2305.19821","created_at":"2026-07-05T06:16:03.649228+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.19821v1","created_at":"2026-07-05T06:16:03.649228+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.19821","created_at":"2026-07-05T06:16:03.649228+00:00"},{"alias_kind":"pith_short_12","alias_value":"DHU4QZFBXADY","created_at":"2026-07-05T06:16:03.649228+00:00"},{"alias_kind":"pith_short_16","alias_value":"DHU4QZFBXADYUWJI","created_at":"2026-07-05T06:16:03.649228+00:00"},{"alias_kind":"pith_short_8","alias_value":"DHU4QZFB","created_at":"2026-07-05T06:16:03.649228+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2501.04513","citing_title":"Improving Image Captioning by Mimicking Human Reformulation Feedback at Inference-time","ref_index":38,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DHU4QZFBXADYUWJIOXJ6W4U2MF","json":"https://pith.science/pith/DHU4QZFBXADYUWJIOXJ6W4U2MF.json","graph_json":"https://pith.science/api/pith-number/DHU4QZFBXADYUWJIOXJ6W4U2MF/graph.json","events_json":"https://pith.science/api/pith-number/DHU4QZFBXADYUWJIOXJ6W4U2MF/events.json","paper":"https://pith.science/paper/DHU4QZFB"},"agent_actions":{"view_html":"https://pith.science/pith/DHU4QZFBXADYUWJIOXJ6W4U2MF","download_json":"https://pith.science/pith/DHU4QZFBXADYUWJIOXJ6W4U2MF.json","view_paper":"https://pith.science/paper/DHU4QZFB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.19821&json=true","fetch_graph":"https://pith.science/api/pith-number/DHU4QZFBXADYUWJIOXJ6W4U2MF/graph.json","fetch_events":"https://pith.science/api/pith-number/DHU4QZFBXADYUWJIOXJ6W4U2MF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DHU4QZFBXADYUWJIOXJ6W4U2MF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DHU4QZFBXADYUWJIOXJ6W4U2MF/action/storage_attestation","attest_author":"https://pith.science/pith/DHU4QZFBXADYUWJIOXJ6W4U2MF/action/author_attestation","sign_citation":"https://pith.science/pith/DHU4QZFBXADYUWJIOXJ6W4U2MF/action/citation_signature","submit_replication":"https://pith.science/pith/DHU4QZFBXADYUWJIOXJ6W4U2MF/action/replication_record"}},"created_at":"2026-07-05T06:16:03.649228+00:00","updated_at":"2026-07-05T06:16:03.649228+00:00"}