{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:NMNRHNFRF5J3BSAHVQTPDEIEG2","short_pith_number":"pith:NMNRHNFR","schema_version":"1.0","canonical_sha256":"6b1b13b4b12f53b0c807ac26f1910436ac2d82a9be5ad9f0dc09463df60cccda","source":{"kind":"arxiv","id":"2507.01297","version":2},"attestation_state":"computed","paper":{"title":"Frustratingly Simple Retrieval Improves Challenging, Reasoning-Intensive Benchmarks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.IR"],"primary_cat":"cs.CL","authors_text":"Michael Duan, Pang Wei Koh, Rulin Shao, Sewon Min, Xinxi Lyu","submitted_at":"2025-07-02T02:35:47Z","abstract_excerpt":"Retrieval-augmented Generation (RAG) has primarily been studied in limited settings, such as factoid question answering; more challenging, reasoning-intensive benchmarks have seen limited success from minimal RAG. In this work, we challenge this prevailing view on established, reasoning-intensive benchmarks: MMLU, MMLU Pro, AGI Eval, GPQA, and MATH. We identify a key missing component in prior work: a usable, web-scale datastore aligned with the breadth of pretraining data. To this end, we introduce CompactDS: a diverse, high-quality, web-scale datastore that achieves high retrieval accuracy a"},"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":"2507.01297","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-07-02T02:35:47Z","cross_cats_sorted":["cs.IR"],"title_canon_sha256":"658fa7752fd3945baa8bd0a2a6c945a1f8b60ad3e7a40b5efb1eac8f59c49b4d","abstract_canon_sha256":"b1a2d8685417495dc9910a0982c8cfb65acee2827a159a43aae4a29e682c9d71"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:32:33.234965Z","signature_b64":"Tb307YnuaZm6f++PVdcsTg6H6GulZMYEJ52lEc5RlkAGlFIxLzv+pffrW0XdXLxxlzgDFErYeWWMVKGswBczCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6b1b13b4b12f53b0c807ac26f1910436ac2d82a9be5ad9f0dc09463df60cccda","last_reissued_at":"2026-07-05T11:32:33.234504Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:32:33.234504Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Frustratingly Simple Retrieval Improves Challenging, Reasoning-Intensive Benchmarks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.IR"],"primary_cat":"cs.CL","authors_text":"Michael Duan, Pang Wei Koh, Rulin Shao, Sewon Min, Xinxi Lyu","submitted_at":"2025-07-02T02:35:47Z","abstract_excerpt":"Retrieval-augmented Generation (RAG) has primarily been studied in limited settings, such as factoid question answering; more challenging, reasoning-intensive benchmarks have seen limited success from minimal RAG. In this work, we challenge this prevailing view on established, reasoning-intensive benchmarks: MMLU, MMLU Pro, AGI Eval, GPQA, and MATH. We identify a key missing component in prior work: a usable, web-scale datastore aligned with the breadth of pretraining data. To this end, we introduce CompactDS: a diverse, high-quality, web-scale datastore that achieves high retrieval accuracy a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.01297","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/2507.01297/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":"2507.01297","created_at":"2026-07-05T11:32:33.234558+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.01297v2","created_at":"2026-07-05T11:32:33.234558+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.01297","created_at":"2026-07-05T11:32:33.234558+00:00"},{"alias_kind":"pith_short_12","alias_value":"NMNRHNFRF5J3","created_at":"2026-07-05T11:32:33.234558+00:00"},{"alias_kind":"pith_short_16","alias_value":"NMNRHNFRF5J3BSAH","created_at":"2026-07-05T11:32:33.234558+00:00"},{"alias_kind":"pith_short_8","alias_value":"NMNRHNFR","created_at":"2026-07-05T11:32:33.234558+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.28344","citing_title":"PIXELRAG: Web Screenshots Beat Text for Retrieval-Augmented Generation","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2604.01348","citing_title":"Procedural Knowledge at Scale Improves Reasoning","ref_index":20,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NMNRHNFRF5J3BSAHVQTPDEIEG2","json":"https://pith.science/pith/NMNRHNFRF5J3BSAHVQTPDEIEG2.json","graph_json":"https://pith.science/api/pith-number/NMNRHNFRF5J3BSAHVQTPDEIEG2/graph.json","events_json":"https://pith.science/api/pith-number/NMNRHNFRF5J3BSAHVQTPDEIEG2/events.json","paper":"https://pith.science/paper/NMNRHNFR"},"agent_actions":{"view_html":"https://pith.science/pith/NMNRHNFRF5J3BSAHVQTPDEIEG2","download_json":"https://pith.science/pith/NMNRHNFRF5J3BSAHVQTPDEIEG2.json","view_paper":"https://pith.science/paper/NMNRHNFR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.01297&json=true","fetch_graph":"https://pith.science/api/pith-number/NMNRHNFRF5J3BSAHVQTPDEIEG2/graph.json","fetch_events":"https://pith.science/api/pith-number/NMNRHNFRF5J3BSAHVQTPDEIEG2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NMNRHNFRF5J3BSAHVQTPDEIEG2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NMNRHNFRF5J3BSAHVQTPDEIEG2/action/storage_attestation","attest_author":"https://pith.science/pith/NMNRHNFRF5J3BSAHVQTPDEIEG2/action/author_attestation","sign_citation":"https://pith.science/pith/NMNRHNFRF5J3BSAHVQTPDEIEG2/action/citation_signature","submit_replication":"https://pith.science/pith/NMNRHNFRF5J3BSAHVQTPDEIEG2/action/replication_record"}},"created_at":"2026-07-05T11:32:33.234558+00:00","updated_at":"2026-07-05T11:32:33.234558+00:00"}