{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:QDYELRVWUJMZYCG6JF43CT4WLR","short_pith_number":"pith:QDYELRVW","schema_version":"1.0","canonical_sha256":"80f045c6b6a2599c08de4979b14f965c5e4c9447ee4b1d55667266143791c9c8","source":{"kind":"arxiv","id":"2305.13534","version":1},"attestation_state":"computed","paper":{"title":"How Language Model Hallucinations Can Snowball","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alisa Liu, Muru Zhang, Noah A. Smith, Ofir Press, William Merrill","submitted_at":"2023-05-22T23:14:28Z","abstract_excerpt":"A major risk of using language models in practical applications is their tendency to hallucinate incorrect statements. Hallucinations are often attributed to knowledge gaps in LMs, but we hypothesize that in some cases, when justifying previously generated hallucinations, LMs output false claims that they can separately recognize as incorrect. We construct three question-answering datasets where ChatGPT and GPT-4 often state an incorrect answer and offer an explanation with at least one incorrect claim. Crucially, we find that ChatGPT and GPT-4 can identify 67% and 87% of their own mistakes, r"},"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":"2305.13534","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-22T23:14:28Z","cross_cats_sorted":[],"title_canon_sha256":"a3bdcfbff37c143182479f820162f1554b04f9e72ade1e02bf667ef75dacc92d","abstract_canon_sha256":"06043fd402ff979d903403b52f8e1cd6cf9cf41758f3813c7e0d14ca92a2b6a7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:12:53.956114Z","signature_b64":"cbvr9cTVxgMx4s9WKEo+It8ez8KddTKkuL1F0c0IbGCiIpMZKP3FOThGiDGKRIrnwMtIneh1rQNpbvq9PruuBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"80f045c6b6a2599c08de4979b14f965c5e4c9447ee4b1d55667266143791c9c8","last_reissued_at":"2026-07-05T06:12:53.955669Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:12:53.955669Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"How Language Model Hallucinations Can Snowball","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alisa Liu, Muru Zhang, Noah A. Smith, Ofir Press, William Merrill","submitted_at":"2023-05-22T23:14:28Z","abstract_excerpt":"A major risk of using language models in practical applications is their tendency to hallucinate incorrect statements. Hallucinations are often attributed to knowledge gaps in LMs, but we hypothesize that in some cases, when justifying previously generated hallucinations, LMs output false claims that they can separately recognize as incorrect. We construct three question-answering datasets where ChatGPT and GPT-4 often state an incorrect answer and offer an explanation with at least one incorrect claim. Crucially, we find that ChatGPT and GPT-4 can identify 67% and 87% of their own mistakes, r"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.13534","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/2305.13534/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":"2305.13534","created_at":"2026-07-05T06:12:53.955728+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.13534v1","created_at":"2026-07-05T06:12:53.955728+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.13534","created_at":"2026-07-05T06:12:53.955728+00:00"},{"alias_kind":"pith_short_12","alias_value":"QDYELRVWUJMZ","created_at":"2026-07-05T06:12:53.955728+00:00"},{"alias_kind":"pith_short_16","alias_value":"QDYELRVWUJMZYCG6","created_at":"2026-07-05T06:12:53.955728+00:00"},{"alias_kind":"pith_short_8","alias_value":"QDYELRVW","created_at":"2026-07-05T06:12:53.955728+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":19,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.08403","citing_title":"Game Theory Driven Multi-Agent Framework Mitigates Language Model Hallucination","ref_index":20,"is_internal_anchor":true},{"citing_arxiv_id":"2606.18060","citing_title":"PseudoBench: Measuring How Agentic Auto-Research Fuels Pseudoscience","ref_index":157,"is_internal_anchor":false},{"citing_arxiv_id":"2606.20662","citing_title":"Confidence Laundering in Agent Systems: Why Uncertainty Needs a Latent Carrier","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2606.27409","citing_title":"Delayed Verification Destabilizes Multi-Agent LLM Belief: Instability Thresholds and Optimal Corrector Placement","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2606.29905","citing_title":"StrucTab: A Structured Optimization Framework for Table Parsing","ref_index":61,"is_internal_anchor":false},{"citing_arxiv_id":"2605.24949","citing_title":"APT-Agent: Automated Penetration Testing using Large Language Models","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2605.27485","citing_title":"Automating Formal Verification with Agent-Guided Tree Search","ref_index":110,"is_internal_anchor":false},{"citing_arxiv_id":"2407.20240","citing_title":"Social and Ethical Risks Posed by General-Purpose LLMs for Settling Newcomers in Canada","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16675","citing_title":"LinAlg-Bench: A Forensic Benchmark Revealing Structural Failure Modes in LLM Mathematical Reasoning","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18930","citing_title":"OEP: Poisoning Self-Evolving LLM Agents via Locally Correct but Non-Transferable Experiences","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2508.18473","citing_title":"Principled Detection of Hallucinations in Large Language Models via Multiple Testing","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2309.11495","citing_title":"Chain-of-Verification Reduces Hallucination in Large Language Models","ref_index":124,"is_internal_anchor":false},{"citing_arxiv_id":"2310.00754","citing_title":"Analyzing and Mitigating Object Hallucination in Large Vision-Language Models","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2404.11584","citing_title":"The Landscape of Emerging AI Agent Architectures for Reasoning, Planning, and Tool Calling: A Survey","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2309.05922","citing_title":"A Survey of Hallucination in Large Foundation Models","ref_index":148,"is_internal_anchor":false},{"citing_arxiv_id":"2602.13595","citing_title":"The Quantization Trap: Breaking Linear Scaling Laws in Multi-Hop Reasoning","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2401.01313","citing_title":"A Comprehensive Survey of Hallucination Mitigation Techniques in Large Language Models","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11217","citing_title":"Leveraging RAG for Training-Free Alignment of LLMs","ref_index":65,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06066","citing_title":"From Hallucination to Structure Snowballing: The Alignment Tax of Constrained Decoding in LLM Reflection","ref_index":14,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QDYELRVWUJMZYCG6JF43CT4WLR","json":"https://pith.science/pith/QDYELRVWUJMZYCG6JF43CT4WLR.json","graph_json":"https://pith.science/api/pith-number/QDYELRVWUJMZYCG6JF43CT4WLR/graph.json","events_json":"https://pith.science/api/pith-number/QDYELRVWUJMZYCG6JF43CT4WLR/events.json","paper":"https://pith.science/paper/QDYELRVW"},"agent_actions":{"view_html":"https://pith.science/pith/QDYELRVWUJMZYCG6JF43CT4WLR","download_json":"https://pith.science/pith/QDYELRVWUJMZYCG6JF43CT4WLR.json","view_paper":"https://pith.science/paper/QDYELRVW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.13534&json=true","fetch_graph":"https://pith.science/api/pith-number/QDYELRVWUJMZYCG6JF43CT4WLR/graph.json","fetch_events":"https://pith.science/api/pith-number/QDYELRVWUJMZYCG6JF43CT4WLR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QDYELRVWUJMZYCG6JF43CT4WLR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QDYELRVWUJMZYCG6JF43CT4WLR/action/storage_attestation","attest_author":"https://pith.science/pith/QDYELRVWUJMZYCG6JF43CT4WLR/action/author_attestation","sign_citation":"https://pith.science/pith/QDYELRVWUJMZYCG6JF43CT4WLR/action/citation_signature","submit_replication":"https://pith.science/pith/QDYELRVWUJMZYCG6JF43CT4WLR/action/replication_record"}},"created_at":"2026-07-05T06:12:53.955728+00:00","updated_at":"2026-07-05T06:12:53.955728+00:00"}