{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:QTCNHMLBJBS3GRRISCYGYTXCG3","short_pith_number":"pith:QTCNHMLB","schema_version":"1.0","canonical_sha256":"84c4d3b1614865b3462890b06c4ee236c740e5a8133431bef429e504d3c452ca","source":{"kind":"arxiv","id":"2305.01579","version":3},"attestation_state":"computed","paper":{"title":"Why So Gullible? Enhancing the Robustness of Retrieval-Augmented Models against Counterfactual Noise","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Giwon Hong, Jeonghwan Kim, Joyce Jiyoung Whang, Junmo Kang, Sung-Hyon Myaeng","submitted_at":"2023-05-02T16:28:10Z","abstract_excerpt":"Most existing retrieval-augmented language models (LMs) assume a naive dichotomy within a retrieved document set: query-relevance and irrelevance. Our work investigates a more challenging scenario in which even the \"relevant\" documents may contain misleading or incorrect information, causing conflict among the retrieved documents and thereby negatively influencing model decisions as noise. We observe that existing LMs are highly brittle to the presence of conflicting information in both the fine-tuning and in-context few-shot learning scenarios. We propose approaches for handling knowledge con"},"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":"2305.01579","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-05-02T16:28:10Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"2c44f5628ff034c1049e620d8e4c569370dac6f170ea4fc1e984fa35e3be8755","abstract_canon_sha256":"0403622b1fefa3e276ed48b090843519283d1bcc069cfb496501af74f6677363"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:29:15.036846Z","signature_b64":"PHYlqrq3PlEFZAybJtAGHNKq9Ok9grpfToiNCcdFQe4cE8QwPK9fppoB3ci/u2W0NNXcL8Nf1mo0YdSIQsF/Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"84c4d3b1614865b3462890b06c4ee236c740e5a8133431bef429e504d3c452ca","last_reissued_at":"2026-07-05T08:29:15.036320Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:29:15.036320Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Why So Gullible? Enhancing the Robustness of Retrieval-Augmented Models against Counterfactual Noise","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Giwon Hong, Jeonghwan Kim, Joyce Jiyoung Whang, Junmo Kang, Sung-Hyon Myaeng","submitted_at":"2023-05-02T16:28:10Z","abstract_excerpt":"Most existing retrieval-augmented language models (LMs) assume a naive dichotomy within a retrieved document set: query-relevance and irrelevance. Our work investigates a more challenging scenario in which even the \"relevant\" documents may contain misleading or incorrect information, causing conflict among the retrieved documents and thereby negatively influencing model decisions as noise. We observe that existing LMs are highly brittle to the presence of conflicting information in both the fine-tuning and in-context few-shot learning scenarios. We propose approaches for handling knowledge con"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.01579","kind":"arxiv","version":3},"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/2305.01579/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":"2305.01579","created_at":"2026-07-05T08:29:15.036382+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.01579v3","created_at":"2026-07-05T08:29:15.036382+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.01579","created_at":"2026-07-05T08:29:15.036382+00:00"},{"alias_kind":"pith_short_12","alias_value":"QTCNHMLBJBS3","created_at":"2026-07-05T08:29:15.036382+00:00"},{"alias_kind":"pith_short_16","alias_value":"QTCNHMLBJBS3GRRI","created_at":"2026-07-05T08:29:15.036382+00:00"},{"alias_kind":"pith_short_8","alias_value":"QTCNHMLB","created_at":"2026-07-05T08:29:15.036382+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2409.10102","citing_title":"Trustworthiness in Retrieval-Augmented Generation Systems: A Survey","ref_index":68,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QTCNHMLBJBS3GRRISCYGYTXCG3","json":"https://pith.science/pith/QTCNHMLBJBS3GRRISCYGYTXCG3.json","graph_json":"https://pith.science/api/pith-number/QTCNHMLBJBS3GRRISCYGYTXCG3/graph.json","events_json":"https://pith.science/api/pith-number/QTCNHMLBJBS3GRRISCYGYTXCG3/events.json","paper":"https://pith.science/paper/QTCNHMLB"},"agent_actions":{"view_html":"https://pith.science/pith/QTCNHMLBJBS3GRRISCYGYTXCG3","download_json":"https://pith.science/pith/QTCNHMLBJBS3GRRISCYGYTXCG3.json","view_paper":"https://pith.science/paper/QTCNHMLB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.01579&json=true","fetch_graph":"https://pith.science/api/pith-number/QTCNHMLBJBS3GRRISCYGYTXCG3/graph.json","fetch_events":"https://pith.science/api/pith-number/QTCNHMLBJBS3GRRISCYGYTXCG3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QTCNHMLBJBS3GRRISCYGYTXCG3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QTCNHMLBJBS3GRRISCYGYTXCG3/action/storage_attestation","attest_author":"https://pith.science/pith/QTCNHMLBJBS3GRRISCYGYTXCG3/action/author_attestation","sign_citation":"https://pith.science/pith/QTCNHMLBJBS3GRRISCYGYTXCG3/action/citation_signature","submit_replication":"https://pith.science/pith/QTCNHMLBJBS3GRRISCYGYTXCG3/action/replication_record"}},"created_at":"2026-07-05T08:29:15.036382+00:00","updated_at":"2026-07-05T08:29:15.036382+00:00"}