{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:DOTXF6VW4FT4UXKTGHRWCDJW5F","short_pith_number":"pith:DOTXF6VW","schema_version":"1.0","canonical_sha256":"1ba772fab6e167ca5d5331e3610d36e97b1929b3c9ff877160e118afff4e278c","source":{"kind":"arxiv","id":"2607.22643","version":1},"attestation_state":"computed","paper":{"title":"Reason Before You Retrieve: Agentic Planning for Multi-modal RAG","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.AI","authors_text":"Minhao Cheng, Shir Simon, Tianyu Yang, Xiangliang Zhang, Zhenzhen Li","submitted_at":"2026-06-24T02:40:49Z","abstract_excerpt":"Multimodal retrieval-augmented generation (mRAG) aims to answer image-text queries with external knowledge, but most existing systems still retrieve directly from raw multimodal input over a flat evidence space. This design often struggles with two key challenges: the retrieval target is under-specified because the question intent must be grounded to the correct visual referent, and the search space is weakly structured, forcing semantically distinct evidence to compete in a single global ranking step. We propose MM-R2, a multimodal agentic retrieval framework that reasons before retrieval by "},"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":"2607.22643","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-24T02:40:49Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"01adc8454bd37317fd0924bf1d47334a0c70de05240a8bd52168a04d7240d5c7","abstract_canon_sha256":"bb367773337c38d73734ac4c805884d0cc0c9c5eb38503de6cf2ef2eda4be8fc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-28T00:21:46.852475Z","signature_b64":"h3zlSw3E5G+6evpbwOSyjmy8X+YnA+H5pyQjJt+lCpqyq9rLFXGYlmNu143Y6jN1tD99Hb2LhPFHH+ZmUMTDAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1ba772fab6e167ca5d5331e3610d36e97b1929b3c9ff877160e118afff4e278c","last_reissued_at":"2026-07-28T00:21:46.851547Z","signature_status":"signed_v1","first_computed_at":"2026-07-28T00:21:46.851547Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Reason Before You Retrieve: Agentic Planning for Multi-modal RAG","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.AI","authors_text":"Minhao Cheng, Shir Simon, Tianyu Yang, Xiangliang Zhang, Zhenzhen Li","submitted_at":"2026-06-24T02:40:49Z","abstract_excerpt":"Multimodal retrieval-augmented generation (mRAG) aims to answer image-text queries with external knowledge, but most existing systems still retrieve directly from raw multimodal input over a flat evidence space. This design often struggles with two key challenges: the retrieval target is under-specified because the question intent must be grounded to the correct visual referent, and the search space is weakly structured, forcing semantically distinct evidence to compete in a single global ranking step. We propose MM-R2, a multimodal agentic retrieval framework that reasons before retrieval by "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.22643","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/2607.22643/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":"2607.22643","created_at":"2026-07-28T00:21:46.851975+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.22643v1","created_at":"2026-07-28T00:21:46.851975+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.22643","created_at":"2026-07-28T00:21:46.851975+00:00"},{"alias_kind":"pith_short_12","alias_value":"DOTXF6VW4FT4","created_at":"2026-07-28T00:21:46.851975+00:00"},{"alias_kind":"pith_short_16","alias_value":"DOTXF6VW4FT4UXKT","created_at":"2026-07-28T00:21:46.851975+00:00"},{"alias_kind":"pith_short_8","alias_value":"DOTXF6VW","created_at":"2026-07-28T00:21:46.851975+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/DOTXF6VW4FT4UXKTGHRWCDJW5F","json":"https://pith.science/pith/DOTXF6VW4FT4UXKTGHRWCDJW5F.json","graph_json":"https://pith.science/api/pith-number/DOTXF6VW4FT4UXKTGHRWCDJW5F/graph.json","events_json":"https://pith.science/api/pith-number/DOTXF6VW4FT4UXKTGHRWCDJW5F/events.json","paper":"https://pith.science/paper/DOTXF6VW"},"agent_actions":{"view_html":"https://pith.science/pith/DOTXF6VW4FT4UXKTGHRWCDJW5F","download_json":"https://pith.science/pith/DOTXF6VW4FT4UXKTGHRWCDJW5F.json","view_paper":"https://pith.science/paper/DOTXF6VW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.22643&json=true","fetch_graph":"https://pith.science/api/pith-number/DOTXF6VW4FT4UXKTGHRWCDJW5F/graph.json","fetch_events":"https://pith.science/api/pith-number/DOTXF6VW4FT4UXKTGHRWCDJW5F/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DOTXF6VW4FT4UXKTGHRWCDJW5F/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DOTXF6VW4FT4UXKTGHRWCDJW5F/action/storage_attestation","attest_author":"https://pith.science/pith/DOTXF6VW4FT4UXKTGHRWCDJW5F/action/author_attestation","sign_citation":"https://pith.science/pith/DOTXF6VW4FT4UXKTGHRWCDJW5F/action/citation_signature","submit_replication":"https://pith.science/pith/DOTXF6VW4FT4UXKTGHRWCDJW5F/action/replication_record"}},"created_at":"2026-07-28T00:21:46.851975+00:00","updated_at":"2026-07-28T00:21:46.851975+00:00"}