{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:3P475OIBYF2DCIOSCDE3QGQBYK","short_pith_number":"pith:3P475OIB","schema_version":"1.0","canonical_sha256":"dbf9feb901c1743121d210c9b81a01c285eb7398b7e356eb41697e0b436b0da6","source":{"kind":"arxiv","id":"2205.03685","version":1},"attestation_state":"computed","paper":{"title":"Better Retrieval May Not Lead to Better Question Answering","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ashish Sabharwal, Mihai Surdeanu, Steven Bethard, Tushar Khot, Zhengzhong Liang","submitted_at":"2022-05-07T16:59:38Z","abstract_excerpt":"Considerable progress has been made recently in open-domain question answering (QA) problems, which require Information Retrieval (IR) and Reading Comprehension (RC). A popular approach to improve the system's performance is to improve the quality of the retrieved context from the IR stage. In this work we show that for StrategyQA, a challenging open-domain QA dataset that requires multi-hop reasoning, this common approach is surprisingly ineffective -- improving the quality of the retrieved context hardly improves the system's performance. We further analyze the system's behavior to identify "},"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":"2205.03685","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-05-07T16:59:38Z","cross_cats_sorted":[],"title_canon_sha256":"f774eb68d1e0da8281efd29cb1f589a96dd630cbfdde4143d72466b22be9ab58","abstract_canon_sha256":"c5433fe3955020181dbab0f72c8f84ed4e9d53080b5f364672d7ec4d21b637d5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:21:17.203134Z","signature_b64":"zKxeEC+rDIZFNLAyE94/ahEDuKdDz+GPsbfoMRvqRwFf+21JAmbkzy0iu536VnuLZOwlDWtexUBX/tlHaZIiBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dbf9feb901c1743121d210c9b81a01c285eb7398b7e356eb41697e0b436b0da6","last_reissued_at":"2026-07-05T04:21:17.202713Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:21:17.202713Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Better Retrieval May Not Lead to Better Question Answering","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ashish Sabharwal, Mihai Surdeanu, Steven Bethard, Tushar Khot, Zhengzhong Liang","submitted_at":"2022-05-07T16:59:38Z","abstract_excerpt":"Considerable progress has been made recently in open-domain question answering (QA) problems, which require Information Retrieval (IR) and Reading Comprehension (RC). A popular approach to improve the system's performance is to improve the quality of the retrieved context from the IR stage. In this work we show that for StrategyQA, a challenging open-domain QA dataset that requires multi-hop reasoning, this common approach is surprisingly ineffective -- improving the quality of the retrieved context hardly improves the system's performance. We further analyze the system's behavior to identify "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2205.03685","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/2205.03685/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":"2205.03685","created_at":"2026-07-05T04:21:17.202773+00:00"},{"alias_kind":"arxiv_version","alias_value":"2205.03685v1","created_at":"2026-07-05T04:21:17.202773+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2205.03685","created_at":"2026-07-05T04:21:17.202773+00:00"},{"alias_kind":"pith_short_12","alias_value":"3P475OIBYF2D","created_at":"2026-07-05T04:21:17.202773+00:00"},{"alias_kind":"pith_short_16","alias_value":"3P475OIBYF2DCIOS","created_at":"2026-07-05T04:21:17.202773+00:00"},{"alias_kind":"pith_short_8","alias_value":"3P475OIB","created_at":"2026-07-05T04:21:17.202773+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2411.16985","citing_title":"Teaching Smaller Language Models To Generalise To Unseen Compositional Questions (Full Thesis)","ref_index":96,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3P475OIBYF2DCIOSCDE3QGQBYK","json":"https://pith.science/pith/3P475OIBYF2DCIOSCDE3QGQBYK.json","graph_json":"https://pith.science/api/pith-number/3P475OIBYF2DCIOSCDE3QGQBYK/graph.json","events_json":"https://pith.science/api/pith-number/3P475OIBYF2DCIOSCDE3QGQBYK/events.json","paper":"https://pith.science/paper/3P475OIB"},"agent_actions":{"view_html":"https://pith.science/pith/3P475OIBYF2DCIOSCDE3QGQBYK","download_json":"https://pith.science/pith/3P475OIBYF2DCIOSCDE3QGQBYK.json","view_paper":"https://pith.science/paper/3P475OIB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2205.03685&json=true","fetch_graph":"https://pith.science/api/pith-number/3P475OIBYF2DCIOSCDE3QGQBYK/graph.json","fetch_events":"https://pith.science/api/pith-number/3P475OIBYF2DCIOSCDE3QGQBYK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3P475OIBYF2DCIOSCDE3QGQBYK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3P475OIBYF2DCIOSCDE3QGQBYK/action/storage_attestation","attest_author":"https://pith.science/pith/3P475OIBYF2DCIOSCDE3QGQBYK/action/author_attestation","sign_citation":"https://pith.science/pith/3P475OIBYF2DCIOSCDE3QGQBYK/action/citation_signature","submit_replication":"https://pith.science/pith/3P475OIBYF2DCIOSCDE3QGQBYK/action/replication_record"}},"created_at":"2026-07-05T04:21:17.202773+00:00","updated_at":"2026-07-05T04:21:17.202773+00:00"}