{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:U2MNZ5MY2R4EB2XMAMSTXYHG3Z","short_pith_number":"pith:U2MNZ5MY","schema_version":"1.0","canonical_sha256":"a698dcf598d47840eaec03253be0e6de6768dde596694939eb61c5e5f179b67e","source":{"kind":"arxiv","id":"2602.04853","version":2},"attestation_state":"computed","paper":{"title":"Decomposed Prompting Does Not Fix Knowledge Gaps, But Helps Models Say \"I Don't Know\"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Dan Roth, Dhruv Madhwal, Lyuxin David Zhang, Tomer Wolfson, Vivek Gupta","submitted_at":"2026-02-04T18:39:58Z","abstract_excerpt":"Large language models often struggle to recognize their knowledge limits in closed-book question answering, leading to confident hallucinations. While decomposed prompting is typically used to improve accuracy, we investigate its impact on reliability. We evaluate three task-equivalent prompting regimes: Direct, Assistive, and Incremental, across different model scales and multi-hop QA benchmarks. We find that although accuracy gains from decomposition diminish in frontier models, disagreements between prompting regimes remain highly indicative of potential errors. Because factual knowledge is"},"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":"2602.04853","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-02-04T18:39:58Z","cross_cats_sorted":[],"title_canon_sha256":"88c0da02e2580274a616d441b823362083bf184ef6fe1b8f61ed0e2aa0831092","abstract_canon_sha256":"ae7a9f4aeea78b01a8253b0137a272b43d873d62392e6a373d9bee0ad5412475"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-07T02:18:34.184642Z","signature_b64":"KjGVCEqwkVWLF6oPlrf7geZqiaq+axowNgtu/YHUGB8nJTvCdLtjKykJlxe3asA03o4n4ubIbIwk6QHNfBGtCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a698dcf598d47840eaec03253be0e6de6768dde596694939eb61c5e5f179b67e","last_reissued_at":"2026-07-07T02:18:34.183544Z","signature_status":"signed_v1","first_computed_at":"2026-07-07T02:18:34.183544Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Decomposed Prompting Does Not Fix Knowledge Gaps, But Helps Models Say \"I Don't Know\"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Dan Roth, Dhruv Madhwal, Lyuxin David Zhang, Tomer Wolfson, Vivek Gupta","submitted_at":"2026-02-04T18:39:58Z","abstract_excerpt":"Large language models often struggle to recognize their knowledge limits in closed-book question answering, leading to confident hallucinations. While decomposed prompting is typically used to improve accuracy, we investigate its impact on reliability. We evaluate three task-equivalent prompting regimes: Direct, Assistive, and Incremental, across different model scales and multi-hop QA benchmarks. We find that although accuracy gains from decomposition diminish in frontier models, disagreements between prompting regimes remain highly indicative of potential errors. Because factual knowledge is"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.04853","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/2602.04853/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":"2602.04853","created_at":"2026-07-07T02:18:34.183680+00:00"},{"alias_kind":"arxiv_version","alias_value":"2602.04853v2","created_at":"2026-07-07T02:18:34.183680+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.04853","created_at":"2026-07-07T02:18:34.183680+00:00"},{"alias_kind":"pith_short_12","alias_value":"U2MNZ5MY2R4E","created_at":"2026-07-07T02:18:34.183680+00:00"},{"alias_kind":"pith_short_16","alias_value":"U2MNZ5MY2R4EB2XM","created_at":"2026-07-07T02:18:34.183680+00:00"},{"alias_kind":"pith_short_8","alias_value":"U2MNZ5MY","created_at":"2026-07-07T02:18:34.183680+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/U2MNZ5MY2R4EB2XMAMSTXYHG3Z","json":"https://pith.science/pith/U2MNZ5MY2R4EB2XMAMSTXYHG3Z.json","graph_json":"https://pith.science/api/pith-number/U2MNZ5MY2R4EB2XMAMSTXYHG3Z/graph.json","events_json":"https://pith.science/api/pith-number/U2MNZ5MY2R4EB2XMAMSTXYHG3Z/events.json","paper":"https://pith.science/paper/U2MNZ5MY"},"agent_actions":{"view_html":"https://pith.science/pith/U2MNZ5MY2R4EB2XMAMSTXYHG3Z","download_json":"https://pith.science/pith/U2MNZ5MY2R4EB2XMAMSTXYHG3Z.json","view_paper":"https://pith.science/paper/U2MNZ5MY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2602.04853&json=true","fetch_graph":"https://pith.science/api/pith-number/U2MNZ5MY2R4EB2XMAMSTXYHG3Z/graph.json","fetch_events":"https://pith.science/api/pith-number/U2MNZ5MY2R4EB2XMAMSTXYHG3Z/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/U2MNZ5MY2R4EB2XMAMSTXYHG3Z/action/timestamp_anchor","attest_storage":"https://pith.science/pith/U2MNZ5MY2R4EB2XMAMSTXYHG3Z/action/storage_attestation","attest_author":"https://pith.science/pith/U2MNZ5MY2R4EB2XMAMSTXYHG3Z/action/author_attestation","sign_citation":"https://pith.science/pith/U2MNZ5MY2R4EB2XMAMSTXYHG3Z/action/citation_signature","submit_replication":"https://pith.science/pith/U2MNZ5MY2R4EB2XMAMSTXYHG3Z/action/replication_record"}},"created_at":"2026-07-07T02:18:34.183680+00:00","updated_at":"2026-07-07T02:18:34.183680+00:00"}