{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:4VUDM7MC43KLYHJQJUO2O3TZTS","short_pith_number":"pith:4VUDM7MC","schema_version":"1.0","canonical_sha256":"e568367d82e6d4bc1d304d1da76e799c8e90373ff6765815e7b425c271dec569","source":{"kind":"arxiv","id":"2410.16166","version":1},"attestation_state":"computed","paper":{"title":"Beyond Filtering: Adaptive Image-Text Quality Enhancement for MLLM Pretraining","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Bingning Wang, Han Huang, Haoyu Lu, Liang Wang, Qiang Liu, Shu Wu, Weipeng Chen, Yuqi Huo, Zijia Zhao","submitted_at":"2024-10-21T16:32:41Z","abstract_excerpt":"Multimodal large language models (MLLMs) have made significant strides by integrating visual and textual modalities. A critical factor in training MLLMs is the quality of image-text pairs within multimodal pretraining datasets. However, $\\textit {de facto}$ filter-based data quality enhancement paradigms often discard a substantial portion of high-quality image data due to inadequate semantic alignment between images and texts, leading to inefficiencies in data utilization and scalability. In this paper, we propose the Adaptive Image-Text Quality Enhancer (AITQE), a model that dynamically asse"},"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":"2410.16166","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-10-21T16:32:41Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"41e353d58ed266476589896cc54fb41acb1ae4e4c09b3f646e685ea516b83516","abstract_canon_sha256":"d33e773b522d66c0ed0e93101167da5c914515b815a558d948eeedd4c0d7608c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:23:30.768200Z","signature_b64":"gwJ/P7yPaRc1opyDFz1cHJs/xtBjzco4oK+E4qIzRt6P3Dgw3XVNDxAiUo4YXIdawv3ddX8zR+sikaI5bB8pAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e568367d82e6d4bc1d304d1da76e799c8e90373ff6765815e7b425c271dec569","last_reissued_at":"2026-07-05T09:23:30.767798Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:23:30.767798Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Beyond Filtering: Adaptive Image-Text Quality Enhancement for MLLM Pretraining","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Bingning Wang, Han Huang, Haoyu Lu, Liang Wang, Qiang Liu, Shu Wu, Weipeng Chen, Yuqi Huo, Zijia Zhao","submitted_at":"2024-10-21T16:32:41Z","abstract_excerpt":"Multimodal large language models (MLLMs) have made significant strides by integrating visual and textual modalities. A critical factor in training MLLMs is the quality of image-text pairs within multimodal pretraining datasets. However, $\\textit {de facto}$ filter-based data quality enhancement paradigms often discard a substantial portion of high-quality image data due to inadequate semantic alignment between images and texts, leading to inefficiencies in data utilization and scalability. In this paper, we propose the Adaptive Image-Text Quality Enhancer (AITQE), a model that dynamically asse"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.16166","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/2410.16166/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":"2410.16166","created_at":"2026-07-05T09:23:30.767856+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.16166v1","created_at":"2026-07-05T09:23:30.767856+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.16166","created_at":"2026-07-05T09:23:30.767856+00:00"},{"alias_kind":"pith_short_12","alias_value":"4VUDM7MC43KL","created_at":"2026-07-05T09:23:30.767856+00:00"},{"alias_kind":"pith_short_16","alias_value":"4VUDM7MC43KLYHJQ","created_at":"2026-07-05T09:23:30.767856+00:00"},{"alias_kind":"pith_short_8","alias_value":"4VUDM7MC","created_at":"2026-07-05T09:23:30.767856+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/4VUDM7MC43KLYHJQJUO2O3TZTS","json":"https://pith.science/pith/4VUDM7MC43KLYHJQJUO2O3TZTS.json","graph_json":"https://pith.science/api/pith-number/4VUDM7MC43KLYHJQJUO2O3TZTS/graph.json","events_json":"https://pith.science/api/pith-number/4VUDM7MC43KLYHJQJUO2O3TZTS/events.json","paper":"https://pith.science/paper/4VUDM7MC"},"agent_actions":{"view_html":"https://pith.science/pith/4VUDM7MC43KLYHJQJUO2O3TZTS","download_json":"https://pith.science/pith/4VUDM7MC43KLYHJQJUO2O3TZTS.json","view_paper":"https://pith.science/paper/4VUDM7MC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.16166&json=true","fetch_graph":"https://pith.science/api/pith-number/4VUDM7MC43KLYHJQJUO2O3TZTS/graph.json","fetch_events":"https://pith.science/api/pith-number/4VUDM7MC43KLYHJQJUO2O3TZTS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4VUDM7MC43KLYHJQJUO2O3TZTS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4VUDM7MC43KLYHJQJUO2O3TZTS/action/storage_attestation","attest_author":"https://pith.science/pith/4VUDM7MC43KLYHJQJUO2O3TZTS/action/author_attestation","sign_citation":"https://pith.science/pith/4VUDM7MC43KLYHJQJUO2O3TZTS/action/citation_signature","submit_replication":"https://pith.science/pith/4VUDM7MC43KLYHJQJUO2O3TZTS/action/replication_record"}},"created_at":"2026-07-05T09:23:30.767856+00:00","updated_at":"2026-07-05T09:23:30.767856+00:00"}