{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:NJYD64TDXYS4HIE2O3GVPMNUMR","short_pith_number":"pith:NJYD64TD","schema_version":"1.0","canonical_sha256":"6a703f7263be25c3a09a76cd57b1b46457641b7588b606226b225ba6d031d5a0","source":{"kind":"arxiv","id":"2404.07060","version":1},"attestation_state":"computed","paper":{"title":"Groundedness in Retrieval-augmented Long-form Generation: An Empirical Study","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Alessandro Stolfo","submitted_at":"2024-04-10T14:50:10Z","abstract_excerpt":"We present an empirical study of groundedness in long-form question answering (LFQA) by retrieval-augmented large language models (LLMs). In particular, we evaluate whether every generated sentence is grounded in the retrieved documents or the model's pre-training data. Across 3 datasets and 4 model families, our findings reveal that a significant fraction of generated sentences are consistently ungrounded, even when those sentences contain correct ground-truth answers. Additionally, we examine the impacts of factors such as model size, decoding strategy, and instruction tuning on groundedness"},"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":"2404.07060","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-10T14:50:10Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"207db90850eb6206aba1934487a44c89c46745e12b1ea01a044e22a1f0d28823","abstract_canon_sha256":"2917e842d5874a681f71419b3392867d2a1909e5f98a3dda504d15f107203639"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:06:34.565998Z","signature_b64":"wmFLmxkl7V9sX9uEij/beGlD4gQtemwcxJSDcHsPbl1DtzKUISekIW0sC3z6JWWZfO9G6cTzUSiukSNi4OznCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6a703f7263be25c3a09a76cd57b1b46457641b7588b606226b225ba6d031d5a0","last_reissued_at":"2026-07-05T08:06:34.565529Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:06:34.565529Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Groundedness in Retrieval-augmented Long-form Generation: An Empirical Study","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Alessandro Stolfo","submitted_at":"2024-04-10T14:50:10Z","abstract_excerpt":"We present an empirical study of groundedness in long-form question answering (LFQA) by retrieval-augmented large language models (LLMs). In particular, we evaluate whether every generated sentence is grounded in the retrieved documents or the model's pre-training data. Across 3 datasets and 4 model families, our findings reveal that a significant fraction of generated sentences are consistently ungrounded, even when those sentences contain correct ground-truth answers. Additionally, we examine the impacts of factors such as model size, decoding strategy, and instruction tuning on groundedness"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.07060","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/2404.07060/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":"2404.07060","created_at":"2026-07-05T08:06:34.565582+00:00"},{"alias_kind":"arxiv_version","alias_value":"2404.07060v1","created_at":"2026-07-05T08:06:34.565582+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.07060","created_at":"2026-07-05T08:06:34.565582+00:00"},{"alias_kind":"pith_short_12","alias_value":"NJYD64TDXYS4","created_at":"2026-07-05T08:06:34.565582+00:00"},{"alias_kind":"pith_short_16","alias_value":"NJYD64TDXYS4HIE2","created_at":"2026-07-05T08:06:34.565582+00:00"},{"alias_kind":"pith_short_8","alias_value":"NJYD64TD","created_at":"2026-07-05T08:06:34.565582+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.09349","citing_title":"Deceptive Grounding: Entity Attribution Failure in Clinical Retrieval-Augmented Generation","ref_index":28,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NJYD64TDXYS4HIE2O3GVPMNUMR","json":"https://pith.science/pith/NJYD64TDXYS4HIE2O3GVPMNUMR.json","graph_json":"https://pith.science/api/pith-number/NJYD64TDXYS4HIE2O3GVPMNUMR/graph.json","events_json":"https://pith.science/api/pith-number/NJYD64TDXYS4HIE2O3GVPMNUMR/events.json","paper":"https://pith.science/paper/NJYD64TD"},"agent_actions":{"view_html":"https://pith.science/pith/NJYD64TDXYS4HIE2O3GVPMNUMR","download_json":"https://pith.science/pith/NJYD64TDXYS4HIE2O3GVPMNUMR.json","view_paper":"https://pith.science/paper/NJYD64TD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2404.07060&json=true","fetch_graph":"https://pith.science/api/pith-number/NJYD64TDXYS4HIE2O3GVPMNUMR/graph.json","fetch_events":"https://pith.science/api/pith-number/NJYD64TDXYS4HIE2O3GVPMNUMR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NJYD64TDXYS4HIE2O3GVPMNUMR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NJYD64TDXYS4HIE2O3GVPMNUMR/action/storage_attestation","attest_author":"https://pith.science/pith/NJYD64TDXYS4HIE2O3GVPMNUMR/action/author_attestation","sign_citation":"https://pith.science/pith/NJYD64TDXYS4HIE2O3GVPMNUMR/action/citation_signature","submit_replication":"https://pith.science/pith/NJYD64TDXYS4HIE2O3GVPMNUMR/action/replication_record"}},"created_at":"2026-07-05T08:06:34.565582+00:00","updated_at":"2026-07-05T08:06:34.565582+00:00"}