{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:75BWMB76RI6I7Q2U2OPAAK5FOH","short_pith_number":"pith:75BWMB76","schema_version":"1.0","canonical_sha256":"ff436607fe8a3c8fc354d39e002ba571dae640e9ceb8df229ab31cdfc9fd9746","source":{"kind":"arxiv","id":"2005.00661","version":1},"attestation_state":"computed","paper":{"title":"On Faithfulness and Factuality in Abstractive Summarization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bernd Bohnet, Joshua Maynez, Ryan McDonald, Shashi Narayan","submitted_at":"2020-05-02T00:09:16Z","abstract_excerpt":"It is well known that the standard likelihood training and approximate decoding objectives in neural text generation models lead to less human-like responses for open-ended tasks such as language modeling and story generation. In this paper we have analyzed limitations of these models for abstractive document summarization and found that these models are highly prone to hallucinate content that is unfaithful to the input document. We conducted a large scale human evaluation of several neural abstractive summarization systems to better understand the types of hallucinations they produce. Our hu"},"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":"2005.00661","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-05-02T00:09:16Z","cross_cats_sorted":[],"title_canon_sha256":"05e5f6b1c5076911a73f24286f31e4a14b321454cec5927743a74c554795ec67","abstract_canon_sha256":"437a8ee8216234de0fc1918cf6ae86e9a84c7d1cc9cc86c95a660a79e30d8045"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:59:44.908558Z","signature_b64":"jMoGmAMAyeThNFzsWp9eN8UcvybSjwGklMqXmHwdsOO3I7SDY6NT+JaJ2sBDHx8DWomsvpjstZPPk6wRZbEEAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ff436607fe8a3c8fc354d39e002ba571dae640e9ceb8df229ab31cdfc9fd9746","last_reissued_at":"2026-07-05T00:59:44.908086Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:59:44.908086Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"On Faithfulness and Factuality in Abstractive Summarization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bernd Bohnet, Joshua Maynez, Ryan McDonald, Shashi Narayan","submitted_at":"2020-05-02T00:09:16Z","abstract_excerpt":"It is well known that the standard likelihood training and approximate decoding objectives in neural text generation models lead to less human-like responses for open-ended tasks such as language modeling and story generation. In this paper we have analyzed limitations of these models for abstractive document summarization and found that these models are highly prone to hallucinate content that is unfaithful to the input document. We conducted a large scale human evaluation of several neural abstractive summarization systems to better understand the types of hallucinations they produce. Our hu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2005.00661","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/2005.00661/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":"2005.00661","created_at":"2026-07-05T00:59:44.908143+00:00"},{"alias_kind":"arxiv_version","alias_value":"2005.00661v1","created_at":"2026-07-05T00:59:44.908143+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2005.00661","created_at":"2026-07-05T00:59:44.908143+00:00"},{"alias_kind":"pith_short_12","alias_value":"75BWMB76RI6I","created_at":"2026-07-05T00:59:44.908143+00:00"},{"alias_kind":"pith_short_16","alias_value":"75BWMB76RI6I7Q2U","created_at":"2026-07-05T00:59:44.908143+00:00"},{"alias_kind":"pith_short_8","alias_value":"75BWMB76","created_at":"2026-07-05T00:59:44.908143+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":25,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.25651","citing_title":"MedGuards: Multi-Agent System for Reliable Medical Error Detection and Correction","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2606.25476","citing_title":"A Red Teaming Framework for Large Language Models: A Case Study on Faithfulness Evaluation","ref_index":62,"is_internal_anchor":false},{"citing_arxiv_id":"2606.22610","citing_title":"PaperClaw: Harnessing Agents for Autonomous Research and Human-in-the-Loop Refinement","ref_index":80,"is_internal_anchor":false},{"citing_arxiv_id":"2606.19591","citing_title":"A BART-based approach with hierarchical strategy for Vietnamese abstractive multi-document summarization","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2606.13216","citing_title":"Layer-Resolved Optimal Transport for Hallucination Detection in NMT and Abstractive Summarization","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2606.08158","citing_title":"Constrained Paraphrase Consistency for LLM Hallucination Detection","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2606.08157","citing_title":"Cross Paraphrastic Invariance Learning for Hallucination Detection","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16953","citing_title":"How do Humans Process AI-generated Hallucination Contents: a Neuroimaging Study","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2606.25651","citing_title":"MedGuards: Multi-Agent System for Reliable Medical Error Detection and Correction","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2605.28264","citing_title":"Entropy Distribution as a Fingerprint for Hallucinations in Generative Models","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2605.29463","citing_title":"Honest Lying: Understanding Memory Confabulation in Reflexive Agents","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2606.26396","citing_title":"At the Edge of Understanding: Sparse Autoencoders Trace The Limits of Transformer Generalization","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2204.06745","citing_title":"GPT-NeoX-20B: An Open-Source Autoregressive Language Model","ref_index":61,"is_internal_anchor":false},{"citing_arxiv_id":"2312.11805","citing_title":"Gemini: A Family of Highly Capable Multimodal Models","ref_index":61,"is_internal_anchor":false},{"citing_arxiv_id":"2412.16720","citing_title":"OpenAI o1 System Card","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2505.06907","citing_title":"A Survey on Foundation Models for Personalized Federated Intelligence","ref_index":239,"is_internal_anchor":false},{"citing_arxiv_id":"2505.16120","citing_title":"LLM-Powered AI Agent Systems and Their Applications in Industry","ref_index":53,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16953","citing_title":"How do Humans Process AI-generated Hallucination Contents: a Neuroimaging Study","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2309.11495","citing_title":"Chain-of-Verification Reduces Hallucination in Large Language Models","ref_index":126,"is_internal_anchor":false},{"citing_arxiv_id":"2511.00739","citing_title":"Towards Understanding, Analyzing, and Optimizing Agentic AI Execution: A CPU-Centric Perspective","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2212.03827","citing_title":"Discovering Latent Knowledge in Language Models Without Supervision","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2604.26197","citing_title":"Hierarchical Long-Term Semantic Memory for LinkedIn's Hiring Agent","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2305.20050","citing_title":"Let's Verify Step by Step","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2604.16686","citing_title":"No-Worse Context-Aware Decoding: Preventing Neutral Regression in Context-Conditioned Generation","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2604.20131","citing_title":"Whose Story Gets Told? Positionality and Bias in LLM Summaries of Life Narratives","ref_index":146,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/75BWMB76RI6I7Q2U2OPAAK5FOH","json":"https://pith.science/pith/75BWMB76RI6I7Q2U2OPAAK5FOH.json","graph_json":"https://pith.science/api/pith-number/75BWMB76RI6I7Q2U2OPAAK5FOH/graph.json","events_json":"https://pith.science/api/pith-number/75BWMB76RI6I7Q2U2OPAAK5FOH/events.json","paper":"https://pith.science/paper/75BWMB76"},"agent_actions":{"view_html":"https://pith.science/pith/75BWMB76RI6I7Q2U2OPAAK5FOH","download_json":"https://pith.science/pith/75BWMB76RI6I7Q2U2OPAAK5FOH.json","view_paper":"https://pith.science/paper/75BWMB76","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2005.00661&json=true","fetch_graph":"https://pith.science/api/pith-number/75BWMB76RI6I7Q2U2OPAAK5FOH/graph.json","fetch_events":"https://pith.science/api/pith-number/75BWMB76RI6I7Q2U2OPAAK5FOH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/75BWMB76RI6I7Q2U2OPAAK5FOH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/75BWMB76RI6I7Q2U2OPAAK5FOH/action/storage_attestation","attest_author":"https://pith.science/pith/75BWMB76RI6I7Q2U2OPAAK5FOH/action/author_attestation","sign_citation":"https://pith.science/pith/75BWMB76RI6I7Q2U2OPAAK5FOH/action/citation_signature","submit_replication":"https://pith.science/pith/75BWMB76RI6I7Q2U2OPAAK5FOH/action/replication_record"}},"created_at":"2026-07-05T00:59:44.908143+00:00","updated_at":"2026-07-05T00:59:44.908143+00:00"}