{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:HXX4ILSECS32BB47LAS3FCVSBA","short_pith_number":"pith:HXX4ILSE","schema_version":"1.0","canonical_sha256":"3defc42e4414b7a0879f5825b28ab2081a3d665fa23724b4d6465f6e7b0e3a6f","source":{"kind":"arxiv","id":"2210.10860","version":1},"attestation_state":"computed","paper":{"title":"Two-Turn Debate Doesn't Help Humans Answer Hard Reading Comprehension Questions","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alicia Parrish, Amanpreet Singh Saimbhi, Harsh Trivedi, Jason Phang, Nikita Nangia, Samuel R. Bowman, Vishakh Padmakumar","submitted_at":"2022-10-19T19:48:50Z","abstract_excerpt":"The use of language-model-based question-answering systems to aid humans in completing difficult tasks is limited, in part, by the unreliability of the text these systems generate. Using hard multiple-choice reading comprehension questions as a testbed, we assess whether presenting humans with arguments for two competing answer options, where one is correct and the other is incorrect, allows human judges to perform more accurately, even when one of the arguments is unreliable and deceptive. If this is helpful, we may be able to increase our justified trust in language-model-based systems by as"},"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":"2210.10860","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-10-19T19:48:50Z","cross_cats_sorted":[],"title_canon_sha256":"146381a61e654c8084c61a0a038c753e5e2cb381703cddbbfe3cd3abc83c231a","abstract_canon_sha256":"dd082b9f4dbd8a42ea2fcfccfaa9c38d29c46829f048f5875239ceb8a6e41266"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:08:32.710894Z","signature_b64":"QlZl+18qDPHBsbDad9d+zo5cUNhX189tBW+BweANVmLzGscx5f8M4m6UM4uULrNVkVyS7zj36QizDgow8lYxAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3defc42e4414b7a0879f5825b28ab2081a3d665fa23724b4d6465f6e7b0e3a6f","last_reissued_at":"2026-07-05T05:08:32.710347Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:08:32.710347Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Two-Turn Debate Doesn't Help Humans Answer Hard Reading Comprehension Questions","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alicia Parrish, Amanpreet Singh Saimbhi, Harsh Trivedi, Jason Phang, Nikita Nangia, Samuel R. Bowman, Vishakh Padmakumar","submitted_at":"2022-10-19T19:48:50Z","abstract_excerpt":"The use of language-model-based question-answering systems to aid humans in completing difficult tasks is limited, in part, by the unreliability of the text these systems generate. Using hard multiple-choice reading comprehension questions as a testbed, we assess whether presenting humans with arguments for two competing answer options, where one is correct and the other is incorrect, allows human judges to perform more accurately, even when one of the arguments is unreliable and deceptive. If this is helpful, we may be able to increase our justified trust in language-model-based systems by as"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2210.10860","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/2210.10860/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":"2210.10860","created_at":"2026-07-05T05:08:32.710418+00:00"},{"alias_kind":"arxiv_version","alias_value":"2210.10860v1","created_at":"2026-07-05T05:08:32.710418+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2210.10860","created_at":"2026-07-05T05:08:32.710418+00:00"},{"alias_kind":"pith_short_12","alias_value":"HXX4ILSECS32","created_at":"2026-07-05T05:08:32.710418+00:00"},{"alias_kind":"pith_short_16","alias_value":"HXX4ILSECS32BB47","created_at":"2026-07-05T05:08:32.710418+00:00"},{"alias_kind":"pith_short_8","alias_value":"HXX4ILSE","created_at":"2026-07-05T05:08:32.710418+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.25133","citing_title":"Trust but Verify: Prover-Verifier Deliberation for Selective LLM Prediction","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2211.03540","citing_title":"Measuring Progress on Scalable Oversight for Large Language Models","ref_index":58,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HXX4ILSECS32BB47LAS3FCVSBA","json":"https://pith.science/pith/HXX4ILSECS32BB47LAS3FCVSBA.json","graph_json":"https://pith.science/api/pith-number/HXX4ILSECS32BB47LAS3FCVSBA/graph.json","events_json":"https://pith.science/api/pith-number/HXX4ILSECS32BB47LAS3FCVSBA/events.json","paper":"https://pith.science/paper/HXX4ILSE"},"agent_actions":{"view_html":"https://pith.science/pith/HXX4ILSECS32BB47LAS3FCVSBA","download_json":"https://pith.science/pith/HXX4ILSECS32BB47LAS3FCVSBA.json","view_paper":"https://pith.science/paper/HXX4ILSE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2210.10860&json=true","fetch_graph":"https://pith.science/api/pith-number/HXX4ILSECS32BB47LAS3FCVSBA/graph.json","fetch_events":"https://pith.science/api/pith-number/HXX4ILSECS32BB47LAS3FCVSBA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HXX4ILSECS32BB47LAS3FCVSBA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HXX4ILSECS32BB47LAS3FCVSBA/action/storage_attestation","attest_author":"https://pith.science/pith/HXX4ILSECS32BB47LAS3FCVSBA/action/author_attestation","sign_citation":"https://pith.science/pith/HXX4ILSECS32BB47LAS3FCVSBA/action/citation_signature","submit_replication":"https://pith.science/pith/HXX4ILSECS32BB47LAS3FCVSBA/action/replication_record"}},"created_at":"2026-07-05T05:08:32.710418+00:00","updated_at":"2026-07-05T05:08:32.710418+00:00"}