{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:SRA4W4BUINKS7NCLIJEVMVN65B","short_pith_number":"pith:SRA4W4BU","schema_version":"1.0","canonical_sha256":"9441cb703443552fb44b42495655bee87bfd3d2c4ed66afcef43c36274c4af53","source":{"kind":"arxiv","id":"2505.19707","version":1},"attestation_state":"computed","paper":{"title":"MLLM-Guided VLM Fine-Tuning with Joint Inference for Zero-Shot Composed Image Retrieval","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.IR"],"primary_cat":"cs.CV","authors_text":"Dacheng Tao, Jingyi Liao, Li Shen, Rong-Cheng Tu, Xiao Luo, Yingjie Wang, Zhao Jin","submitted_at":"2025-05-26T08:56:59Z","abstract_excerpt":"Existing Zero-Shot Composed Image Retrieval (ZS-CIR) methods typically train adapters that convert reference images into pseudo-text tokens, which are concatenated with the modifying text and processed by frozen text encoders in pretrained VLMs or LLMs. While this design leverages the strengths of large pretrained models, it only supervises the adapter to produce encoder-compatible tokens that loosely preserve visual semantics. Crucially, it does not directly optimize the composed query representation to capture the full intent of the composition or to align with the target semantics, thereby "},"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":"2505.19707","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-05-26T08:56:59Z","cross_cats_sorted":["cs.IR"],"title_canon_sha256":"cf162abc99c3e73d1e7994b28ad27276bd729cd1141a81070c60a508cd814557","abstract_canon_sha256":"c8967a28dd8f0120044297f6688fde614cab3a06cc8a59ef0eb96c5c110ef312"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:09:34.902883Z","signature_b64":"1yN3J87sfhTLlBMKvE0qXwGECCrVuF/2IjfiTzXQ8oiMCrc4yEjpwLlii8GlCSMQiR412XWlahUuKMPkyaTyBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9441cb703443552fb44b42495655bee87bfd3d2c4ed66afcef43c36274c4af53","last_reissued_at":"2026-07-05T11:09:34.902394Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:09:34.902394Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MLLM-Guided VLM Fine-Tuning with Joint Inference for Zero-Shot Composed Image Retrieval","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.IR"],"primary_cat":"cs.CV","authors_text":"Dacheng Tao, Jingyi Liao, Li Shen, Rong-Cheng Tu, Xiao Luo, Yingjie Wang, Zhao Jin","submitted_at":"2025-05-26T08:56:59Z","abstract_excerpt":"Existing Zero-Shot Composed Image Retrieval (ZS-CIR) methods typically train adapters that convert reference images into pseudo-text tokens, which are concatenated with the modifying text and processed by frozen text encoders in pretrained VLMs or LLMs. While this design leverages the strengths of large pretrained models, it only supervises the adapter to produce encoder-compatible tokens that loosely preserve visual semantics. Crucially, it does not directly optimize the composed query representation to capture the full intent of the composition or to align with the target semantics, thereby "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.19707","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/2505.19707/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":"2505.19707","created_at":"2026-07-05T11:09:34.902448+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.19707v1","created_at":"2026-07-05T11:09:34.902448+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.19707","created_at":"2026-07-05T11:09:34.902448+00:00"},{"alias_kind":"pith_short_12","alias_value":"SRA4W4BUINKS","created_at":"2026-07-05T11:09:34.902448+00:00"},{"alias_kind":"pith_short_16","alias_value":"SRA4W4BUINKS7NCL","created_at":"2026-07-05T11:09:34.902448+00:00"},{"alias_kind":"pith_short_8","alias_value":"SRA4W4BU","created_at":"2026-07-05T11:09:34.902448+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/SRA4W4BUINKS7NCLIJEVMVN65B","json":"https://pith.science/pith/SRA4W4BUINKS7NCLIJEVMVN65B.json","graph_json":"https://pith.science/api/pith-number/SRA4W4BUINKS7NCLIJEVMVN65B/graph.json","events_json":"https://pith.science/api/pith-number/SRA4W4BUINKS7NCLIJEVMVN65B/events.json","paper":"https://pith.science/paper/SRA4W4BU"},"agent_actions":{"view_html":"https://pith.science/pith/SRA4W4BUINKS7NCLIJEVMVN65B","download_json":"https://pith.science/pith/SRA4W4BUINKS7NCLIJEVMVN65B.json","view_paper":"https://pith.science/paper/SRA4W4BU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.19707&json=true","fetch_graph":"https://pith.science/api/pith-number/SRA4W4BUINKS7NCLIJEVMVN65B/graph.json","fetch_events":"https://pith.science/api/pith-number/SRA4W4BUINKS7NCLIJEVMVN65B/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SRA4W4BUINKS7NCLIJEVMVN65B/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SRA4W4BUINKS7NCLIJEVMVN65B/action/storage_attestation","attest_author":"https://pith.science/pith/SRA4W4BUINKS7NCLIJEVMVN65B/action/author_attestation","sign_citation":"https://pith.science/pith/SRA4W4BUINKS7NCLIJEVMVN65B/action/citation_signature","submit_replication":"https://pith.science/pith/SRA4W4BUINKS7NCLIJEVMVN65B/action/replication_record"}},"created_at":"2026-07-05T11:09:34.902448+00:00","updated_at":"2026-07-05T11:09:34.902448+00:00"}