{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:6BMKK4RBWPEZI7N45EEG7LEC4S","short_pith_number":"pith:6BMKK4RB","schema_version":"1.0","canonical_sha256":"f058a57221b3c9947dbce9086fac82e4b8023d1e89a4f68aaf0a5f2f1788f8b7","source":{"kind":"arxiv","id":"2607.00447","version":1},"attestation_state":"computed","paper":{"title":"Understanding Why Language Models Hallucinate: Testing Reasoning Against Priors","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Haoyue Bai, Jiawei Zhang, Robert Nowak, Shashank Muralidhar Bharadwaj, Siyang Cao, Xi Ding, Xuhan Tong, Yangfan Hu","submitted_at":"2026-07-01T05:02:43Z","abstract_excerpt":"Large language models often produce hallucinated answers that violate prompt-level constraints. A key diagnostic question is whether these failures reflect missing knowledge, or whether the model has the relevant information but follows the wrong inference path. We study this phenomenon as inference misalignment: a mismatch between the answer supported by the prompt and the answer favored by statistically salient latent associations. We formalize this view with a latent key-task model, in which pretraining-frequency imbalance can cause a shortcut path to dominate the constraint-sensitive path "},"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":"2607.00447","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-07-01T05:02:43Z","cross_cats_sorted":[],"title_canon_sha256":"6bdc86c7fac56076cc8996fb0e6cb14fc161f5b8f54f9b8b2cfcd5f1da16af20","abstract_canon_sha256":"03a19497d5b4fe8a707eb9b42d2d6f453ad3295b744a9f83de9c79a63f9492ca"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-02T01:17:43.634547Z","signature_b64":"viOxxNh7uYfKB1VlHUD72V5cQ7ofgxAn32NEDge6X8LzIfad+ZirTRq/UJMHbRco7Rc0GM7f/xkK4t54A3TGAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f058a57221b3c9947dbce9086fac82e4b8023d1e89a4f68aaf0a5f2f1788f8b7","last_reissued_at":"2026-07-02T01:17:43.634154Z","signature_status":"signed_v1","first_computed_at":"2026-07-02T01:17:43.634154Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Understanding Why Language Models Hallucinate: Testing Reasoning Against Priors","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Haoyue Bai, Jiawei Zhang, Robert Nowak, Shashank Muralidhar Bharadwaj, Siyang Cao, Xi Ding, Xuhan Tong, Yangfan Hu","submitted_at":"2026-07-01T05:02:43Z","abstract_excerpt":"Large language models often produce hallucinated answers that violate prompt-level constraints. A key diagnostic question is whether these failures reflect missing knowledge, or whether the model has the relevant information but follows the wrong inference path. We study this phenomenon as inference misalignment: a mismatch between the answer supported by the prompt and the answer favored by statistically salient latent associations. We formalize this view with a latent key-task model, in which pretraining-frequency imbalance can cause a shortcut path to dominate the constraint-sensitive path "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.00447","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/2607.00447/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":"2607.00447","created_at":"2026-07-02T01:17:43.634209+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.00447v1","created_at":"2026-07-02T01:17:43.634209+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.00447","created_at":"2026-07-02T01:17:43.634209+00:00"},{"alias_kind":"pith_short_12","alias_value":"6BMKK4RBWPEZ","created_at":"2026-07-02T01:17:43.634209+00:00"},{"alias_kind":"pith_short_16","alias_value":"6BMKK4RBWPEZI7N4","created_at":"2026-07-02T01:17:43.634209+00:00"},{"alias_kind":"pith_short_8","alias_value":"6BMKK4RB","created_at":"2026-07-02T01:17:43.634209+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/6BMKK4RBWPEZI7N45EEG7LEC4S","json":"https://pith.science/pith/6BMKK4RBWPEZI7N45EEG7LEC4S.json","graph_json":"https://pith.science/api/pith-number/6BMKK4RBWPEZI7N45EEG7LEC4S/graph.json","events_json":"https://pith.science/api/pith-number/6BMKK4RBWPEZI7N45EEG7LEC4S/events.json","paper":"https://pith.science/paper/6BMKK4RB"},"agent_actions":{"view_html":"https://pith.science/pith/6BMKK4RBWPEZI7N45EEG7LEC4S","download_json":"https://pith.science/pith/6BMKK4RBWPEZI7N45EEG7LEC4S.json","view_paper":"https://pith.science/paper/6BMKK4RB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.00447&json=true","fetch_graph":"https://pith.science/api/pith-number/6BMKK4RBWPEZI7N45EEG7LEC4S/graph.json","fetch_events":"https://pith.science/api/pith-number/6BMKK4RBWPEZI7N45EEG7LEC4S/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6BMKK4RBWPEZI7N45EEG7LEC4S/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6BMKK4RBWPEZI7N45EEG7LEC4S/action/storage_attestation","attest_author":"https://pith.science/pith/6BMKK4RBWPEZI7N45EEG7LEC4S/action/author_attestation","sign_citation":"https://pith.science/pith/6BMKK4RBWPEZI7N45EEG7LEC4S/action/citation_signature","submit_replication":"https://pith.science/pith/6BMKK4RBWPEZI7N45EEG7LEC4S/action/replication_record"}},"created_at":"2026-07-02T01:17:43.634209+00:00","updated_at":"2026-07-02T01:17:43.634209+00:00"}