{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:JCDTTYJMGGCR24ELTUSKQO2PL7","short_pith_number":"pith:JCDTTYJM","schema_version":"1.0","canonical_sha256":"488739e12c31851d708b9d24a83b4f5fea8204991ab3a3646f5991a34c2df4b7","source":{"kind":"arxiv","id":"2406.02265","version":3},"attestation_state":"computed","paper":{"title":"Understanding Retrieval Robustness for Retrieval-Augmented Image Captioning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Desmond Elliott, Jiaang Li, Raphael Tang, Rita Ramos, Wenyan Li","submitted_at":"2024-06-04T12:41:54Z","abstract_excerpt":"Recent advances in retrieval-augmented models for image captioning highlight the benefit of retrieving related captions for efficient, lightweight models with strong domain-transfer capabilities. While these models demonstrate the success of retrieval augmentation, retrieval models are still far from perfect in practice: the retrieved information can sometimes mislead the model, resulting in incorrect generation and worse performance. In this paper, we analyze the robustness of a retrieval-augmented captioning model SmallCap. Our analysis shows that the model is sensitive to tokens that appear"},"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":"2406.02265","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-06-04T12:41:54Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"962d539725e114341168e6cc6a6e5644089b7e23b418a3a2f2db9165dea591e7","abstract_canon_sha256":"f0db282ba7caf77d5bb1cff1357adcde44ac2b71370c53f716da54055de0cb0d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:52:33.658364Z","signature_b64":"6J51U723AAWsMDWRIxmoisFQsOado52dMrBfP6RYEMcYIWT/ZlKBXcE3cOUU7seMV4vGAGAzwUGKhLOgztwdBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"488739e12c31851d708b9d24a83b4f5fea8204991ab3a3646f5991a34c2df4b7","last_reissued_at":"2026-07-05T08:52:33.657871Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:52:33.657871Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Understanding Retrieval Robustness for Retrieval-Augmented Image Captioning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Desmond Elliott, Jiaang Li, Raphael Tang, Rita Ramos, Wenyan Li","submitted_at":"2024-06-04T12:41:54Z","abstract_excerpt":"Recent advances in retrieval-augmented models for image captioning highlight the benefit of retrieving related captions for efficient, lightweight models with strong domain-transfer capabilities. While these models demonstrate the success of retrieval augmentation, retrieval models are still far from perfect in practice: the retrieved information can sometimes mislead the model, resulting in incorrect generation and worse performance. In this paper, we analyze the robustness of a retrieval-augmented captioning model SmallCap. Our analysis shows that the model is sensitive to tokens that appear"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.02265","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/2406.02265/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":"2406.02265","created_at":"2026-07-05T08:52:33.657934+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.02265v3","created_at":"2026-07-05T08:52:33.657934+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.02265","created_at":"2026-07-05T08:52:33.657934+00:00"},{"alias_kind":"pith_short_12","alias_value":"JCDTTYJMGGCR","created_at":"2026-07-05T08:52:33.657934+00:00"},{"alias_kind":"pith_short_16","alias_value":"JCDTTYJMGGCR24EL","created_at":"2026-07-05T08:52:33.657934+00:00"},{"alias_kind":"pith_short_8","alias_value":"JCDTTYJM","created_at":"2026-07-05T08:52:33.657934+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.18553","citing_title":"Hierarchical Multi-Modal Retrieval for Knowledge-Grounded News Image Captioning","ref_index":7,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JCDTTYJMGGCR24ELTUSKQO2PL7","json":"https://pith.science/pith/JCDTTYJMGGCR24ELTUSKQO2PL7.json","graph_json":"https://pith.science/api/pith-number/JCDTTYJMGGCR24ELTUSKQO2PL7/graph.json","events_json":"https://pith.science/api/pith-number/JCDTTYJMGGCR24ELTUSKQO2PL7/events.json","paper":"https://pith.science/paper/JCDTTYJM"},"agent_actions":{"view_html":"https://pith.science/pith/JCDTTYJMGGCR24ELTUSKQO2PL7","download_json":"https://pith.science/pith/JCDTTYJMGGCR24ELTUSKQO2PL7.json","view_paper":"https://pith.science/paper/JCDTTYJM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.02265&json=true","fetch_graph":"https://pith.science/api/pith-number/JCDTTYJMGGCR24ELTUSKQO2PL7/graph.json","fetch_events":"https://pith.science/api/pith-number/JCDTTYJMGGCR24ELTUSKQO2PL7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JCDTTYJMGGCR24ELTUSKQO2PL7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JCDTTYJMGGCR24ELTUSKQO2PL7/action/storage_attestation","attest_author":"https://pith.science/pith/JCDTTYJMGGCR24ELTUSKQO2PL7/action/author_attestation","sign_citation":"https://pith.science/pith/JCDTTYJMGGCR24ELTUSKQO2PL7/action/citation_signature","submit_replication":"https://pith.science/pith/JCDTTYJMGGCR24ELTUSKQO2PL7/action/replication_record"}},"created_at":"2026-07-05T08:52:33.657934+00:00","updated_at":"2026-07-05T08:52:33.657934+00:00"}