{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:QF5VABJKI66E6FV6V3E3TTDIAE","short_pith_number":"pith:QF5VABJK","schema_version":"1.0","canonical_sha256":"817b50052a47bc4f16beaec9b9cc68010d4a2b44e4ba24e8e4c7e41d7044a3e4","source":{"kind":"arxiv","id":"2505.17061","version":3},"attestation_state":"computed","paper":{"title":"Mixture of Decoding: An Attention-Inspired Adaptive Decoding Strategy to Mitigate Hallucinations in Large Vision-Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.CL","authors_text":"Fuzheng Zhang, Junfei Wu, Qiang Liu, Tieniu Tan, Xinlong Chen, Yuanxing Zhang","submitted_at":"2025-05-17T09:44:18Z","abstract_excerpt":"Large Vision-Language Models (LVLMs) have exhibited impressive capabilities across various visual tasks, yet they remain hindered by the persistent challenge of hallucinations. To address this critical issue, we propose Mixture of Decoding (MoD), a novel approach for hallucination mitigation that dynamically adapts decoding strategies by evaluating the correctness of the model's attention on image tokens. Specifically, MoD measures the consistency between outputs generated from the original image tokens and those derived from the model's attended image tokens, to distinguish the correctness af"},"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":"2505.17061","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-17T09:44:18Z","cross_cats_sorted":["cs.AI","cs.CV"],"title_canon_sha256":"251a2a5ef91bf31319a7b046256f543e444959038ac3c26a1c11019ae5babee8","abstract_canon_sha256":"460059afcb1e7ed1f82a335ae04b5314615308ffd71900e0cf3e84f24e02e325"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:18:43.090395Z","signature_b64":"AmVTCkj2d/YRSU11QtEINqotZulHpgMXG7Ag+2+c7LPTKPVZJPriEMhVsnowLZFwq66YmHhc64sgJhJpr3q3AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"817b50052a47bc4f16beaec9b9cc68010d4a2b44e4ba24e8e4c7e41d7044a3e4","last_reissued_at":"2026-07-05T11:18:43.089923Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:18:43.089923Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Mixture of Decoding: An Attention-Inspired Adaptive Decoding Strategy to Mitigate Hallucinations in Large Vision-Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.CL","authors_text":"Fuzheng Zhang, Junfei Wu, Qiang Liu, Tieniu Tan, Xinlong Chen, Yuanxing Zhang","submitted_at":"2025-05-17T09:44:18Z","abstract_excerpt":"Large Vision-Language Models (LVLMs) have exhibited impressive capabilities across various visual tasks, yet they remain hindered by the persistent challenge of hallucinations. To address this critical issue, we propose Mixture of Decoding (MoD), a novel approach for hallucination mitigation that dynamically adapts decoding strategies by evaluating the correctness of the model's attention on image tokens. Specifically, MoD measures the consistency between outputs generated from the original image tokens and those derived from the model's attended image tokens, to distinguish the correctness af"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.17061","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/2505.17061/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":"2505.17061","created_at":"2026-07-05T11:18:43.089977+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.17061v3","created_at":"2026-07-05T11:18:43.089977+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.17061","created_at":"2026-07-05T11:18:43.089977+00:00"},{"alias_kind":"pith_short_12","alias_value":"QF5VABJKI66E","created_at":"2026-07-05T11:18:43.089977+00:00"},{"alias_kind":"pith_short_16","alias_value":"QF5VABJKI66E6FV6","created_at":"2026-07-05T11:18:43.089977+00:00"},{"alias_kind":"pith_short_8","alias_value":"QF5VABJK","created_at":"2026-07-05T11:18:43.089977+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.27596","citing_title":"Dismantling Pathological Shortcuts: A Causal Framework for Faithful LVLM Decoding","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2604.17982","citing_title":"Mitigating Multimodal Hallucination via Phase-wise Self-reward","ref_index":9,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QF5VABJKI66E6FV6V3E3TTDIAE","json":"https://pith.science/pith/QF5VABJKI66E6FV6V3E3TTDIAE.json","graph_json":"https://pith.science/api/pith-number/QF5VABJKI66E6FV6V3E3TTDIAE/graph.json","events_json":"https://pith.science/api/pith-number/QF5VABJKI66E6FV6V3E3TTDIAE/events.json","paper":"https://pith.science/paper/QF5VABJK"},"agent_actions":{"view_html":"https://pith.science/pith/QF5VABJKI66E6FV6V3E3TTDIAE","download_json":"https://pith.science/pith/QF5VABJKI66E6FV6V3E3TTDIAE.json","view_paper":"https://pith.science/paper/QF5VABJK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.17061&json=true","fetch_graph":"https://pith.science/api/pith-number/QF5VABJKI66E6FV6V3E3TTDIAE/graph.json","fetch_events":"https://pith.science/api/pith-number/QF5VABJKI66E6FV6V3E3TTDIAE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QF5VABJKI66E6FV6V3E3TTDIAE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QF5VABJKI66E6FV6V3E3TTDIAE/action/storage_attestation","attest_author":"https://pith.science/pith/QF5VABJKI66E6FV6V3E3TTDIAE/action/author_attestation","sign_citation":"https://pith.science/pith/QF5VABJKI66E6FV6V3E3TTDIAE/action/citation_signature","submit_replication":"https://pith.science/pith/QF5VABJKI66E6FV6V3E3TTDIAE/action/replication_record"}},"created_at":"2026-07-05T11:18:43.089977+00:00","updated_at":"2026-07-05T11:18:43.089977+00:00"}