{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:USWHWPJKTXMXUPFWLW2C6BQ566","short_pith_number":"pith:USWHWPJK","schema_version":"1.0","canonical_sha256":"a4ac7b3d2a9dd97a3cb65db42f061df7b1e661b398323de3ce5ce1752b8a5443","source":{"kind":"arxiv","id":"2504.01883","version":1},"attestation_state":"computed","paper":{"title":"CoRAG: Collaborative Retrieval-Augmented Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.IR","cs.LG"],"primary_cat":"cs.AI","authors_text":"Aashiq Muhamed, Mona Diab, Virginia Smith","submitted_at":"2025-04-02T16:40:43Z","abstract_excerpt":"Retrieval-Augmented Generation (RAG) models excel in knowledge-intensive tasks, especially under few-shot learning constraints. We introduce CoRAG, a framework extending RAG to collaborative settings, where clients jointly train a shared model using a collaborative passage store. To evaluate CoRAG, we introduce CRAB, a benchmark for collaborative homogeneous open-domain question answering. Our experiments demonstrate that CoRAG consistently outperforms both parametric collaborative learning methods and locally trained RAG models in low-resource scenarios. Further analysis reveals the critical "},"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":"2504.01883","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-04-02T16:40:43Z","cross_cats_sorted":["cs.CL","cs.IR","cs.LG"],"title_canon_sha256":"ff254737a798e7b658ad7553a34dee4776c6754ee236020ffa8ea91d20e1ef12","abstract_canon_sha256":"b1965b226f315396e21ac8a611b9373bb0ea29acb21d79d5dc16938b1e4461c8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:43:31.209886Z","signature_b64":"E6201uQMMXCw4vwaaPPWxTkMCAvgEOUDHKkP2JrkIL0UUPxmNM5RNxm51csFjPIjGB6bUM3eG1dIyv+PvklbDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a4ac7b3d2a9dd97a3cb65db42f061df7b1e661b398323de3ce5ce1752b8a5443","last_reissued_at":"2026-07-05T10:43:31.209458Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:43:31.209458Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CoRAG: Collaborative Retrieval-Augmented Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.IR","cs.LG"],"primary_cat":"cs.AI","authors_text":"Aashiq Muhamed, Mona Diab, Virginia Smith","submitted_at":"2025-04-02T16:40:43Z","abstract_excerpt":"Retrieval-Augmented Generation (RAG) models excel in knowledge-intensive tasks, especially under few-shot learning constraints. We introduce CoRAG, a framework extending RAG to collaborative settings, where clients jointly train a shared model using a collaborative passage store. To evaluate CoRAG, we introduce CRAB, a benchmark for collaborative homogeneous open-domain question answering. Our experiments demonstrate that CoRAG consistently outperforms both parametric collaborative learning methods and locally trained RAG models in low-resource scenarios. Further analysis reveals the critical "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.01883","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/2504.01883/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":"2504.01883","created_at":"2026-07-05T10:43:31.209514+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.01883v1","created_at":"2026-07-05T10:43:31.209514+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.01883","created_at":"2026-07-05T10:43:31.209514+00:00"},{"alias_kind":"pith_short_12","alias_value":"USWHWPJKTXMX","created_at":"2026-07-05T10:43:31.209514+00:00"},{"alias_kind":"pith_short_16","alias_value":"USWHWPJKTXMXUPFW","created_at":"2026-07-05T10:43:31.209514+00:00"},{"alias_kind":"pith_short_8","alias_value":"USWHWPJK","created_at":"2026-07-05T10:43:31.209514+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/USWHWPJKTXMXUPFWLW2C6BQ566","json":"https://pith.science/pith/USWHWPJKTXMXUPFWLW2C6BQ566.json","graph_json":"https://pith.science/api/pith-number/USWHWPJKTXMXUPFWLW2C6BQ566/graph.json","events_json":"https://pith.science/api/pith-number/USWHWPJKTXMXUPFWLW2C6BQ566/events.json","paper":"https://pith.science/paper/USWHWPJK"},"agent_actions":{"view_html":"https://pith.science/pith/USWHWPJKTXMXUPFWLW2C6BQ566","download_json":"https://pith.science/pith/USWHWPJKTXMXUPFWLW2C6BQ566.json","view_paper":"https://pith.science/paper/USWHWPJK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.01883&json=true","fetch_graph":"https://pith.science/api/pith-number/USWHWPJKTXMXUPFWLW2C6BQ566/graph.json","fetch_events":"https://pith.science/api/pith-number/USWHWPJKTXMXUPFWLW2C6BQ566/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/USWHWPJKTXMXUPFWLW2C6BQ566/action/timestamp_anchor","attest_storage":"https://pith.science/pith/USWHWPJKTXMXUPFWLW2C6BQ566/action/storage_attestation","attest_author":"https://pith.science/pith/USWHWPJKTXMXUPFWLW2C6BQ566/action/author_attestation","sign_citation":"https://pith.science/pith/USWHWPJKTXMXUPFWLW2C6BQ566/action/citation_signature","submit_replication":"https://pith.science/pith/USWHWPJKTXMXUPFWLW2C6BQ566/action/replication_record"}},"created_at":"2026-07-05T10:43:31.209514+00:00","updated_at":"2026-07-05T10:43:31.209514+00:00"}