{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:HLIUMWDXHFVG7GWECW2TNOQWHH","short_pith_number":"pith:HLIUMWDX","schema_version":"1.0","canonical_sha256":"3ad1465877396a6f9ac415b536ba1639d4f0ef92393cb3ac3f68eb79ff204ecd","source":{"kind":"arxiv","id":"2504.04915","version":1},"attestation_state":"computed","paper":{"title":"Collab-RAG: Boosting Retrieval-Augmented Generation for Complex Question Answering via White-Box and Black-Box LLM Collaboration","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.IR","cs.LG"],"primary_cat":"cs.CL","authors_text":"Carl Yang, Haoyu Wang, Joyce C. Ho, Ran Xu, Wenqi Shi, Yuchen Zhuang, Yue Yu","submitted_at":"2025-04-07T10:52:22Z","abstract_excerpt":"Retrieval-Augmented Generation (RAG) systems often struggle to handle multi-hop question-answering tasks accurately due to irrelevant context retrieval and limited complex reasoning capabilities. We introduce Collab-RAG, a collaborative training framework that leverages mutual enhancement between a white-box small language model (SLM) and a blackbox large language model (LLM) for RAG. Specifically, the SLM decomposes complex queries into simpler sub-questions, thus enhancing the accuracy of the retrieval and facilitating more effective reasoning by the black-box LLM. Concurrently, the black-bo"},"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":"2504.04915","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-07T10:52:22Z","cross_cats_sorted":["cs.AI","cs.IR","cs.LG"],"title_canon_sha256":"c57ab9f1c73d30be28457cb1874d3e7c00a0268e0542d0e8af45edb96c74a57d","abstract_canon_sha256":"c6dc2c012d3742bbd5c6e0541260a9f2a6ef8dc457461378042167348b0791b5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:45:31.499543Z","signature_b64":"yfHpMrPrl4nK2dxfEJPvSpYXdBnNKDYNXsZwF3NOCtpDOPYySM8bCXIiLnDCpj2uqM9SEiaIf66U4ncJi2QDAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3ad1465877396a6f9ac415b536ba1639d4f0ef92393cb3ac3f68eb79ff204ecd","last_reissued_at":"2026-07-05T10:45:31.499144Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:45:31.499144Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Collab-RAG: Boosting Retrieval-Augmented Generation for Complex Question Answering via White-Box and Black-Box LLM Collaboration","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.IR","cs.LG"],"primary_cat":"cs.CL","authors_text":"Carl Yang, Haoyu Wang, Joyce C. Ho, Ran Xu, Wenqi Shi, Yuchen Zhuang, Yue Yu","submitted_at":"2025-04-07T10:52:22Z","abstract_excerpt":"Retrieval-Augmented Generation (RAG) systems often struggle to handle multi-hop question-answering tasks accurately due to irrelevant context retrieval and limited complex reasoning capabilities. We introduce Collab-RAG, a collaborative training framework that leverages mutual enhancement between a white-box small language model (SLM) and a blackbox large language model (LLM) for RAG. Specifically, the SLM decomposes complex queries into simpler sub-questions, thus enhancing the accuracy of the retrieval and facilitating more effective reasoning by the black-box LLM. Concurrently, the black-bo"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.04915","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/2504.04915/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":"2504.04915","created_at":"2026-07-05T10:45:31.499203+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.04915v1","created_at":"2026-07-05T10:45:31.499203+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.04915","created_at":"2026-07-05T10:45:31.499203+00:00"},{"alias_kind":"pith_short_12","alias_value":"HLIUMWDXHFVG","created_at":"2026-07-05T10:45:31.499203+00:00"},{"alias_kind":"pith_short_16","alias_value":"HLIUMWDXHFVG7GWE","created_at":"2026-07-05T10:45:31.499203+00:00"},{"alias_kind":"pith_short_8","alias_value":"HLIUMWDX","created_at":"2026-07-05T10:45:31.499203+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.13221","citing_title":"An Agentic AI Framework with Large Language Models and Chain-of-Thought for UAV-Assisted Logistics Scheduling with Mobile Edge Computing","ref_index":46,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HLIUMWDXHFVG7GWECW2TNOQWHH","json":"https://pith.science/pith/HLIUMWDXHFVG7GWECW2TNOQWHH.json","graph_json":"https://pith.science/api/pith-number/HLIUMWDXHFVG7GWECW2TNOQWHH/graph.json","events_json":"https://pith.science/api/pith-number/HLIUMWDXHFVG7GWECW2TNOQWHH/events.json","paper":"https://pith.science/paper/HLIUMWDX"},"agent_actions":{"view_html":"https://pith.science/pith/HLIUMWDXHFVG7GWECW2TNOQWHH","download_json":"https://pith.science/pith/HLIUMWDXHFVG7GWECW2TNOQWHH.json","view_paper":"https://pith.science/paper/HLIUMWDX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.04915&json=true","fetch_graph":"https://pith.science/api/pith-number/HLIUMWDXHFVG7GWECW2TNOQWHH/graph.json","fetch_events":"https://pith.science/api/pith-number/HLIUMWDXHFVG7GWECW2TNOQWHH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HLIUMWDXHFVG7GWECW2TNOQWHH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HLIUMWDXHFVG7GWECW2TNOQWHH/action/storage_attestation","attest_author":"https://pith.science/pith/HLIUMWDXHFVG7GWECW2TNOQWHH/action/author_attestation","sign_citation":"https://pith.science/pith/HLIUMWDXHFVG7GWECW2TNOQWHH/action/citation_signature","submit_replication":"https://pith.science/pith/HLIUMWDXHFVG7GWECW2TNOQWHH/action/replication_record"}},"created_at":"2026-07-05T10:45:31.499203+00:00","updated_at":"2026-07-05T10:45:31.499203+00:00"}