{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:JYCIZBHTGAWIBBVYDHFIOMGHZF","short_pith_number":"pith:JYCIZBHT","schema_version":"1.0","canonical_sha256":"4e048c84f3302c8086b819ca8730c7c96ebb129f2d88df4051dfdca3f476ffb2","source":{"kind":"arxiv","id":"2409.01658","version":3},"attestation_state":"computed","paper":{"title":"From Yes-Men to Truth-Tellers: Addressing Sycophancy in Large Language Models with Pinpoint Tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Binbin Lin, Deng Cai, Houqiang Li, Jieping Ye, Le Lu, Liang Xie, Wei Chen, Wenxiao Wang, Xinmei Tian, Xu Shen, Yonggang Zhang, Zhen Huang","submitted_at":"2024-09-03T07:01:37Z","abstract_excerpt":"Large Language Models (LLMs) tend to prioritize adherence to user prompts over providing veracious responses, leading to the sycophancy issue. When challenged by users, LLMs tend to admit mistakes and provide inaccurate responses even if they initially provided the correct answer. Recent works propose to employ supervised fine-tuning (SFT) to mitigate the sycophancy issue, while it typically leads to the degeneration of LLMs' general capability. To address the challenge, we propose a novel supervised pinpoint tuning (SPT), where the region-of-interest modules are tuned for a given objective. S"},"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":"2409.01658","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-09-03T07:01:37Z","cross_cats_sorted":[],"title_canon_sha256":"17af85065989b76379779253c5d77b1220da561ff1634cacde408e1d2408c598","abstract_canon_sha256":"a46ba85a5a6e9d1b5c89303935b3515dd332bcc3cfc8c0f0e8ad3e8c16e763be"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:09:47.188460Z","signature_b64":"IKSwJOwkLG4GtXY/uH3LkttGUSiyB8XDT7pIp/yY2B7pxR5/5tgfM2K2Dbnxu3EmCAcFZEgf4HC/XwdaPtq7AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4e048c84f3302c8086b819ca8730c7c96ebb129f2d88df4051dfdca3f476ffb2","last_reissued_at":"2026-07-05T10:09:47.187910Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:09:47.187910Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"From Yes-Men to Truth-Tellers: Addressing Sycophancy in Large Language Models with Pinpoint Tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Binbin Lin, Deng Cai, Houqiang Li, Jieping Ye, Le Lu, Liang Xie, Wei Chen, Wenxiao Wang, Xinmei Tian, Xu Shen, Yonggang Zhang, Zhen Huang","submitted_at":"2024-09-03T07:01:37Z","abstract_excerpt":"Large Language Models (LLMs) tend to prioritize adherence to user prompts over providing veracious responses, leading to the sycophancy issue. When challenged by users, LLMs tend to admit mistakes and provide inaccurate responses even if they initially provided the correct answer. Recent works propose to employ supervised fine-tuning (SFT) to mitigate the sycophancy issue, while it typically leads to the degeneration of LLMs' general capability. To address the challenge, we propose a novel supervised pinpoint tuning (SPT), where the region-of-interest modules are tuned for a given objective. S"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.01658","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/2409.01658/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":"2409.01658","created_at":"2026-07-05T10:09:47.187970+00:00"},{"alias_kind":"arxiv_version","alias_value":"2409.01658v3","created_at":"2026-07-05T10:09:47.187970+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.01658","created_at":"2026-07-05T10:09:47.187970+00:00"},{"alias_kind":"pith_short_12","alias_value":"JYCIZBHTGAWI","created_at":"2026-07-05T10:09:47.187970+00:00"},{"alias_kind":"pith_short_16","alias_value":"JYCIZBHTGAWIBBVY","created_at":"2026-07-05T10:09:47.187970+00:00"},{"alias_kind":"pith_short_8","alias_value":"JYCIZBHT","created_at":"2026-07-05T10:09:47.187970+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":8,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.07916","citing_title":"Persona Cartography: Charting Language Model Personality Traits in Weight Space","ref_index":13,"is_internal_anchor":true},{"citing_arxiv_id":"2607.01071","citing_title":"MemSyco-Bench: Benchmarking Sycophancy in Agent Memory","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2606.10949","citing_title":"Recalling Too Well: Sycophancy Evaluation and Mitigation in Memory-Augmented Models","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2607.01071","citing_title":"MemSyco-Bench: Benchmarking Sycophancy in Agent Memory","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2508.16846","citing_title":"BASIL: Bayesian Assessment of Sycophancy in LLMs","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2603.18740","citing_title":"Measuring and Exploiting Contextual Bias in LLM-Assisted Security Code Review","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2604.24668","citing_title":"The Price of Agreement: Measuring LLM Sycophancy in Agentic Financial Applications","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2604.05279","citing_title":"Pressure, What Pressure? Sycophancy Disentanglement in Language Models via Reward Decomposition","ref_index":3,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JYCIZBHTGAWIBBVYDHFIOMGHZF","json":"https://pith.science/pith/JYCIZBHTGAWIBBVYDHFIOMGHZF.json","graph_json":"https://pith.science/api/pith-number/JYCIZBHTGAWIBBVYDHFIOMGHZF/graph.json","events_json":"https://pith.science/api/pith-number/JYCIZBHTGAWIBBVYDHFIOMGHZF/events.json","paper":"https://pith.science/paper/JYCIZBHT"},"agent_actions":{"view_html":"https://pith.science/pith/JYCIZBHTGAWIBBVYDHFIOMGHZF","download_json":"https://pith.science/pith/JYCIZBHTGAWIBBVYDHFIOMGHZF.json","view_paper":"https://pith.science/paper/JYCIZBHT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2409.01658&json=true","fetch_graph":"https://pith.science/api/pith-number/JYCIZBHTGAWIBBVYDHFIOMGHZF/graph.json","fetch_events":"https://pith.science/api/pith-number/JYCIZBHTGAWIBBVYDHFIOMGHZF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JYCIZBHTGAWIBBVYDHFIOMGHZF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JYCIZBHTGAWIBBVYDHFIOMGHZF/action/storage_attestation","attest_author":"https://pith.science/pith/JYCIZBHTGAWIBBVYDHFIOMGHZF/action/author_attestation","sign_citation":"https://pith.science/pith/JYCIZBHTGAWIBBVYDHFIOMGHZF/action/citation_signature","submit_replication":"https://pith.science/pith/JYCIZBHTGAWIBBVYDHFIOMGHZF/action/replication_record"}},"created_at":"2026-07-05T10:09:47.187970+00:00","updated_at":"2026-07-05T10:09:47.187970+00:00"}