{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:FFZKSTMR2G5MIUBLZGEG2R6P6U","short_pith_number":"pith:FFZKSTMR","schema_version":"1.0","canonical_sha256":"2972a94d91d1bac4502bc9886d47cff5086052b64ce46b316a36ffb3550198e7","source":{"kind":"arxiv","id":"2605.01284","version":2},"attestation_state":"computed","paper":{"title":"Chain of Evidence: Pixel-Level Visual Attribution for Iterative Retrieval-Augmented Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Vision-language models can deliver pixel-level visual evidence chains for iterative retrieval-augmented generation by operating directly on document screenshots.","cross_cats":["cs.AI","cs.CL","cs.IR"],"primary_cat":"cs.CV","authors_text":"Di Liang, Peiyang Liu, Wei Ye, Xi Wang, Ziqiang Cui","submitted_at":"2026-05-02T06:40:42Z","abstract_excerpt":"Iterative Retrieval-Augmented Generation (iRAG) has emerged as a powerful paradigm for answering complex multi-hop questions by progressively retrieving and reasoning over external documents. However, current systems predominantly operate on parsed text, which creates two critical bottlenecks: (1) \\textit{Coarse-grained attribution}, where users are burdened with manually locating evidence within lengthy documents based on vague text-level citations; and (2) \\textit{Visual semantic loss}, where the conversion of visually rich documents (e.g., slides, PDFs with charts) into text discards spatia"},"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":"2605.01284","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-02T06:40:42Z","cross_cats_sorted":["cs.AI","cs.CL","cs.IR"],"title_canon_sha256":"d08b1190e6ccce64ca8ad55466d40ca9103daa69a66afaeaa7dd9130256fe2f7","abstract_canon_sha256":"118bf059e09fbe7927df304ab6d8bb2c863f08dcaba997245484db0b223c2087"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T01:03:31.681437Z","signature_b64":"6/BtKp86jUs0NFzghgZgCTthmcRYz19eFVhpP7QByZMhm43RUMKabxqYyOw+kuw7cIxmLv+XBNR5A4SEHfDbDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2972a94d91d1bac4502bc9886d47cff5086052b64ce46b316a36ffb3550198e7","last_reissued_at":"2026-05-26T01:03:31.680631Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T01:03:31.680631Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Chain of Evidence: Pixel-Level Visual Attribution for Iterative Retrieval-Augmented Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Vision-language models can deliver pixel-level visual evidence chains for iterative retrieval-augmented generation by operating directly on document screenshots.","cross_cats":["cs.AI","cs.CL","cs.IR"],"primary_cat":"cs.CV","authors_text":"Di Liang, Peiyang Liu, Wei Ye, Xi Wang, Ziqiang Cui","submitted_at":"2026-05-02T06:40:42Z","abstract_excerpt":"Iterative Retrieval-Augmented Generation (iRAG) has emerged as a powerful paradigm for answering complex multi-hop questions by progressively retrieving and reasoning over external documents. However, current systems predominantly operate on parsed text, which creates two critical bottlenecks: (1) \\textit{Coarse-grained attribution}, where users are burdened with manually locating evidence within lengthy documents based on vague text-level citations; and (2) \\textit{Visual semantic loss}, where the conversion of visually rich documents (e.g., slides, PDFs with charts) into text discards spatia"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"fine-tuned Qwen3-VL-8B-Instruct achieves robust performance, significantly outperforming text-based baselines in scenarios requiring visual layout understanding, while establishing a retriever-agnostic solution for pixel-level interpretable iRAG.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That vision-language models applied to raw document screenshots can reliably recover spatial logic and layout cues that text conversion discards, without format-specific parsing or additional supervision beyond the fine-tuning described.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"CoE applies vision-language models directly to document screenshots to deliver pixel-level bounding-box attribution for evidence in iterative retrieval-augmented generation, outperforming text baselines on visual-layout tasks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Vision-language models can deliver pixel-level visual evidence chains for iterative retrieval-augmented generation by operating directly on document screenshots.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"aac36e72c896125b8a1411355e904d9260963437b70889c9d8e7cd40bb9a88a5"},"source":{"id":"2605.01284","kind":"arxiv","version":2},"verdict":{"id":"ee448b55-1017-4652-b5e6-87e9ee16626d","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-09T14:42:48.086500Z","strongest_claim":"fine-tuned Qwen3-VL-8B-Instruct achieves robust performance, significantly outperforming text-based baselines in scenarios requiring visual layout understanding, while establishing a retriever-agnostic solution for pixel-level interpretable iRAG.","one_line_summary":"CoE applies vision-language models directly to document screenshots to deliver pixel-level bounding-box attribution for evidence in iterative retrieval-augmented generation, outperforming text baselines on visual-layout tasks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That vision-language models applied to raw document screenshots can reliably recover spatial logic and layout cues that text conversion discards, without format-specific parsing or additional supervision beyond the fine-tuning described.","pith_extraction_headline":"Vision-language models can deliver pixel-level visual evidence chains for iterative retrieval-augmented generation by operating directly on document screenshots."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.01284/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-20T18:35:45.366112Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T17:27:00.327473Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"650733d44fd774141e603140059a529820898b3fa431f48d1df8b89b55a1e878"},"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":"2605.01284","created_at":"2026-05-26T01:03:31.680746+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.01284v2","created_at":"2026-05-26T01:03:31.680746+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.01284","created_at":"2026-05-26T01:03:31.680746+00:00"},{"alias_kind":"pith_short_12","alias_value":"FFZKSTMR2G5M","created_at":"2026-05-26T01:03:31.680746+00:00"},{"alias_kind":"pith_short_16","alias_value":"FFZKSTMR2G5MIUBL","created_at":"2026-05-26T01:03:31.680746+00:00"},{"alias_kind":"pith_short_8","alias_value":"FFZKSTMR","created_at":"2026-05-26T01:03:31.680746+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/FFZKSTMR2G5MIUBLZGEG2R6P6U","json":"https://pith.science/pith/FFZKSTMR2G5MIUBLZGEG2R6P6U.json","graph_json":"https://pith.science/api/pith-number/FFZKSTMR2G5MIUBLZGEG2R6P6U/graph.json","events_json":"https://pith.science/api/pith-number/FFZKSTMR2G5MIUBLZGEG2R6P6U/events.json","paper":"https://pith.science/paper/FFZKSTMR"},"agent_actions":{"view_html":"https://pith.science/pith/FFZKSTMR2G5MIUBLZGEG2R6P6U","download_json":"https://pith.science/pith/FFZKSTMR2G5MIUBLZGEG2R6P6U.json","view_paper":"https://pith.science/paper/FFZKSTMR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.01284&json=true","fetch_graph":"https://pith.science/api/pith-number/FFZKSTMR2G5MIUBLZGEG2R6P6U/graph.json","fetch_events":"https://pith.science/api/pith-number/FFZKSTMR2G5MIUBLZGEG2R6P6U/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FFZKSTMR2G5MIUBLZGEG2R6P6U/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FFZKSTMR2G5MIUBLZGEG2R6P6U/action/storage_attestation","attest_author":"https://pith.science/pith/FFZKSTMR2G5MIUBLZGEG2R6P6U/action/author_attestation","sign_citation":"https://pith.science/pith/FFZKSTMR2G5MIUBLZGEG2R6P6U/action/citation_signature","submit_replication":"https://pith.science/pith/FFZKSTMR2G5MIUBLZGEG2R6P6U/action/replication_record"}},"created_at":"2026-05-26T01:03:31.680746+00:00","updated_at":"2026-05-26T01:03:31.680746+00:00"}