{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:JS6QYCYFT66HFTAM4YMBENSKCW","short_pith_number":"pith:JS6QYCYF","schema_version":"1.0","canonical_sha256":"4cbd0c0b059fbc72cc0ce61812364a15b617339adb9ace51050046e460d659c3","source":{"kind":"arxiv","id":"2407.16833","version":2},"attestation_state":"computed","paper":{"title":"Retrieval Augmented Generation or Long-Context LLMs? A Comprehensive Study and Hybrid Approach","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Cheng Li, Michael Bendersky, Mingyang Zhang, Qiaozhu Mei, Zhuowan Li","submitted_at":"2024-07-23T20:51:52Z","abstract_excerpt":"Retrieval Augmented Generation (RAG) has been a powerful tool for Large Language Models (LLMs) to efficiently process overly lengthy contexts. However, recent LLMs like Gemini-1.5 and GPT-4 show exceptional capabilities to understand long contexts directly. We conduct a comprehensive comparison between RAG and long-context (LC) LLMs, aiming to leverage the strengths of both. We benchmark RAG and LC across various public datasets using three latest LLMs. Results reveal that when resourced sufficiently, LC consistently outperforms RAG in terms of average performance. However, RAG's significantly"},"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":"2407.16833","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-23T20:51:52Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"00fd019048365c671355f78ad7a877b47a3545fed353424f1c131dbe294c82c4","abstract_canon_sha256":"d71359ea61cc967caab1e7cf643db6ed139ebf199dfececfe8b0a1f8325239b0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:21:47.261773Z","signature_b64":"LA0I6GDv60dibV1WWZgm/mGgrs+2Oe584qfbPraL1aP7H7qX32JvrFghrUByLkkP4BB1LMJpd5KMPWzQmqzlDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4cbd0c0b059fbc72cc0ce61812364a15b617339adb9ace51050046e460d659c3","last_reissued_at":"2026-07-05T09:21:47.261148Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:21:47.261148Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Retrieval Augmented Generation or Long-Context LLMs? A Comprehensive Study and Hybrid Approach","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Cheng Li, Michael Bendersky, Mingyang Zhang, Qiaozhu Mei, Zhuowan Li","submitted_at":"2024-07-23T20:51:52Z","abstract_excerpt":"Retrieval Augmented Generation (RAG) has been a powerful tool for Large Language Models (LLMs) to efficiently process overly lengthy contexts. However, recent LLMs like Gemini-1.5 and GPT-4 show exceptional capabilities to understand long contexts directly. We conduct a comprehensive comparison between RAG and long-context (LC) LLMs, aiming to leverage the strengths of both. We benchmark RAG and LC across various public datasets using three latest LLMs. Results reveal that when resourced sufficiently, LC consistently outperforms RAG in terms of average performance. However, RAG's significantly"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.16833","kind":"arxiv","version":2},"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/2407.16833/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":"2407.16833","created_at":"2026-07-05T09:21:47.261205+00:00"},{"alias_kind":"arxiv_version","alias_value":"2407.16833v2","created_at":"2026-07-05T09:21:47.261205+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.16833","created_at":"2026-07-05T09:21:47.261205+00:00"},{"alias_kind":"pith_short_12","alias_value":"JS6QYCYFT66H","created_at":"2026-07-05T09:21:47.261205+00:00"},{"alias_kind":"pith_short_16","alias_value":"JS6QYCYFT66HFTAM","created_at":"2026-07-05T09:21:47.261205+00:00"},{"alias_kind":"pith_short_8","alias_value":"JS6QYCYF","created_at":"2026-07-05T09:21:47.261205+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.11759","citing_title":"Retrieval Is Not Enough: Why Organizational AI Needs Epistemic Infrastructure","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2604.11759","citing_title":"Retrieval Is Not Enough: Why Organizational AI Needs Epistemic Infrastructure","ref_index":23,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JS6QYCYFT66HFTAM4YMBENSKCW","json":"https://pith.science/pith/JS6QYCYFT66HFTAM4YMBENSKCW.json","graph_json":"https://pith.science/api/pith-number/JS6QYCYFT66HFTAM4YMBENSKCW/graph.json","events_json":"https://pith.science/api/pith-number/JS6QYCYFT66HFTAM4YMBENSKCW/events.json","paper":"https://pith.science/paper/JS6QYCYF"},"agent_actions":{"view_html":"https://pith.science/pith/JS6QYCYFT66HFTAM4YMBENSKCW","download_json":"https://pith.science/pith/JS6QYCYFT66HFTAM4YMBENSKCW.json","view_paper":"https://pith.science/paper/JS6QYCYF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2407.16833&json=true","fetch_graph":"https://pith.science/api/pith-number/JS6QYCYFT66HFTAM4YMBENSKCW/graph.json","fetch_events":"https://pith.science/api/pith-number/JS6QYCYFT66HFTAM4YMBENSKCW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JS6QYCYFT66HFTAM4YMBENSKCW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JS6QYCYFT66HFTAM4YMBENSKCW/action/storage_attestation","attest_author":"https://pith.science/pith/JS6QYCYFT66HFTAM4YMBENSKCW/action/author_attestation","sign_citation":"https://pith.science/pith/JS6QYCYFT66HFTAM4YMBENSKCW/action/citation_signature","submit_replication":"https://pith.science/pith/JS6QYCYFT66HFTAM4YMBENSKCW/action/replication_record"}},"created_at":"2026-07-05T09:21:47.261205+00:00","updated_at":"2026-07-05T09:21:47.261205+00:00"}