{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:F45DEJCQZH5XR3ALO2YCDDC5WR","short_pith_number":"pith:F45DEJCQ","schema_version":"1.0","canonical_sha256":"2f3a322450c9fb78ec0b76b0218c5db441167212fbea033b34989d42aab19686","source":{"kind":"arxiv","id":"2410.15926","version":1},"attestation_state":"computed","paper":{"title":"Mitigating Object Hallucination via Concentric Causal Attention","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Ivan Laptev, Shijian Lu, Yiheng Li, Yun Xing","submitted_at":"2024-10-21T11:54:53Z","abstract_excerpt":"Recent Large Vision Language Models (LVLMs) present remarkable zero-shot conversational and reasoning capabilities given multimodal queries. Nevertheless, they suffer from object hallucination, a phenomenon where LVLMs are prone to generate textual responses not factually aligned with image inputs. Our pilot study reveals that object hallucination is closely tied with Rotary Position Encoding (RoPE), a widely adopted positional dependency modeling design in existing LVLMs. Due to the long-term decay in RoPE, LVLMs tend to hallucinate more when relevant visual cues are distant from instruction "},"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":"2410.15926","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-10-21T11:54:53Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"605dafff46a31d07ccbbf0af5527cdd7f61e599dbca428601b6b091f1d126721","abstract_canon_sha256":"83ffef755a1807a3d3f055db81c087f94e8828ff442c7799e4169ce49294f859"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:23:26.739779Z","signature_b64":"zEn9IesOE/h8RrU3j7BmyzEnwcVf7XjhwJaTZz/YZmkvOE3IUKO33jebF1OExMneNIeWq1Mf7Jl8xWgfC8sSCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2f3a322450c9fb78ec0b76b0218c5db441167212fbea033b34989d42aab19686","last_reissued_at":"2026-07-05T09:23:26.739294Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:23:26.739294Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Mitigating Object Hallucination via Concentric Causal Attention","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Ivan Laptev, Shijian Lu, Yiheng Li, Yun Xing","submitted_at":"2024-10-21T11:54:53Z","abstract_excerpt":"Recent Large Vision Language Models (LVLMs) present remarkable zero-shot conversational and reasoning capabilities given multimodal queries. Nevertheless, they suffer from object hallucination, a phenomenon where LVLMs are prone to generate textual responses not factually aligned with image inputs. Our pilot study reveals that object hallucination is closely tied with Rotary Position Encoding (RoPE), a widely adopted positional dependency modeling design in existing LVLMs. Due to the long-term decay in RoPE, LVLMs tend to hallucinate more when relevant visual cues are distant from instruction "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.15926","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/2410.15926/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":"2410.15926","created_at":"2026-07-05T09:23:26.739365+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.15926v1","created_at":"2026-07-05T09:23:26.739365+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.15926","created_at":"2026-07-05T09:23:26.739365+00:00"},{"alias_kind":"pith_short_12","alias_value":"F45DEJCQZH5X","created_at":"2026-07-05T09:23:26.739365+00:00"},{"alias_kind":"pith_short_16","alias_value":"F45DEJCQZH5XR3AL","created_at":"2026-07-05T09:23:26.739365+00:00"},{"alias_kind":"pith_short_8","alias_value":"F45DEJCQ","created_at":"2026-07-05T09:23:26.739365+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.29805","citing_title":"Clearer Sight, Fewer Lies: Oriented Pickup Preference Optimization for Multimodal Hallucination Mitigation","ref_index":72,"is_internal_anchor":false},{"citing_arxiv_id":"2606.29805","citing_title":"Clearer Sight, Fewer Lies: Oriented Pickup Preference Optimization for Multimodal Hallucination Mitigation","ref_index":72,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11559","citing_title":"When Looking Is Not Enough: Visual Attention Structure Reveals Hallucination in MLLMs","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2605.04641","citing_title":"CAST: Mitigating Object Hallucination in Large Vision-Language Models via Caption-Guided Visual Attention Steering","ref_index":56,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/F45DEJCQZH5XR3ALO2YCDDC5WR","json":"https://pith.science/pith/F45DEJCQZH5XR3ALO2YCDDC5WR.json","graph_json":"https://pith.science/api/pith-number/F45DEJCQZH5XR3ALO2YCDDC5WR/graph.json","events_json":"https://pith.science/api/pith-number/F45DEJCQZH5XR3ALO2YCDDC5WR/events.json","paper":"https://pith.science/paper/F45DEJCQ"},"agent_actions":{"view_html":"https://pith.science/pith/F45DEJCQZH5XR3ALO2YCDDC5WR","download_json":"https://pith.science/pith/F45DEJCQZH5XR3ALO2YCDDC5WR.json","view_paper":"https://pith.science/paper/F45DEJCQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.15926&json=true","fetch_graph":"https://pith.science/api/pith-number/F45DEJCQZH5XR3ALO2YCDDC5WR/graph.json","fetch_events":"https://pith.science/api/pith-number/F45DEJCQZH5XR3ALO2YCDDC5WR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/F45DEJCQZH5XR3ALO2YCDDC5WR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/F45DEJCQZH5XR3ALO2YCDDC5WR/action/storage_attestation","attest_author":"https://pith.science/pith/F45DEJCQZH5XR3ALO2YCDDC5WR/action/author_attestation","sign_citation":"https://pith.science/pith/F45DEJCQZH5XR3ALO2YCDDC5WR/action/citation_signature","submit_replication":"https://pith.science/pith/F45DEJCQZH5XR3ALO2YCDDC5WR/action/replication_record"}},"created_at":"2026-07-05T09:23:26.739365+00:00","updated_at":"2026-07-05T09:23:26.739365+00:00"}