{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:3KYWDOKOMWZFNSDNZHBLZYKTMZ","short_pith_number":"pith:3KYWDOKO","schema_version":"1.0","canonical_sha256":"dab161b94e65b256c86dc9c2bce1536657907d0b1f8ddeff51ec5a9fa607a302","source":{"kind":"arxiv","id":"2403.00816","version":3},"attestation_state":"computed","paper":{"title":"Read and Think: An Efficient Step-wise Multimodal Language Model for Document Understanding and Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.IR","authors_text":"Jinxu Zhang","submitted_at":"2024-02-26T01:17:50Z","abstract_excerpt":"Understanding the contents of multimodal documents is essential to accurately extract relevant evidence and use it for reasoning. Existing document understanding models tend to generate answers with a single word or phrase directly, ignoring the source document's evidence and lacking interpretability. In this work, we address the lack of step-wise capabilities through data augmentation and extension. Specifically, We use Multi-modal Large Language Models (MLLMs), which have strong visual understanding and reasoning abilities, as data generators to generate step-wise question-and-answer pairs f"},"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":"2403.00816","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2024-02-26T01:17:50Z","cross_cats_sorted":["cs.AI","cs.CV"],"title_canon_sha256":"c06378824bac7d60940bc840d0106c74bf80a588417e37a288e7900dcb099521","abstract_canon_sha256":"6173ba3298a470e9cdeb28af0ddff170fd895fa1e5a045d89ab4182a7fb7b4f4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:55:08.112900Z","signature_b64":"PNahJ4yg8Z/KFzN9jG640djhT33Z9Y6D/1Bs2Abf3Vn1zoNxBVFvva65NMUXqdu1wKXGaHh8f8J4eseyqepVAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dab161b94e65b256c86dc9c2bce1536657907d0b1f8ddeff51ec5a9fa607a302","last_reissued_at":"2026-07-05T08:55:08.112490Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:55:08.112490Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Read and Think: An Efficient Step-wise Multimodal Language Model for Document Understanding and Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.IR","authors_text":"Jinxu Zhang","submitted_at":"2024-02-26T01:17:50Z","abstract_excerpt":"Understanding the contents of multimodal documents is essential to accurately extract relevant evidence and use it for reasoning. Existing document understanding models tend to generate answers with a single word or phrase directly, ignoring the source document's evidence and lacking interpretability. In this work, we address the lack of step-wise capabilities through data augmentation and extension. Specifically, We use Multi-modal Large Language Models (MLLMs), which have strong visual understanding and reasoning abilities, as data generators to generate step-wise question-and-answer pairs f"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.00816","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/2403.00816/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":"2403.00816","created_at":"2026-07-05T08:55:08.112547+00:00"},{"alias_kind":"arxiv_version","alias_value":"2403.00816v3","created_at":"2026-07-05T08:55:08.112547+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.00816","created_at":"2026-07-05T08:55:08.112547+00:00"},{"alias_kind":"pith_short_12","alias_value":"3KYWDOKOMWZF","created_at":"2026-07-05T08:55:08.112547+00:00"},{"alias_kind":"pith_short_16","alias_value":"3KYWDOKOMWZFNSDN","created_at":"2026-07-05T08:55:08.112547+00:00"},{"alias_kind":"pith_short_8","alias_value":"3KYWDOKO","created_at":"2026-07-05T08:55:08.112547+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.23813","citing_title":"ShredBench: Evaluating the Semantic Reasoning Capabilities of Multimodal LLMs in Document Reconstruction","ref_index":28,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3KYWDOKOMWZFNSDNZHBLZYKTMZ","json":"https://pith.science/pith/3KYWDOKOMWZFNSDNZHBLZYKTMZ.json","graph_json":"https://pith.science/api/pith-number/3KYWDOKOMWZFNSDNZHBLZYKTMZ/graph.json","events_json":"https://pith.science/api/pith-number/3KYWDOKOMWZFNSDNZHBLZYKTMZ/events.json","paper":"https://pith.science/paper/3KYWDOKO"},"agent_actions":{"view_html":"https://pith.science/pith/3KYWDOKOMWZFNSDNZHBLZYKTMZ","download_json":"https://pith.science/pith/3KYWDOKOMWZFNSDNZHBLZYKTMZ.json","view_paper":"https://pith.science/paper/3KYWDOKO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2403.00816&json=true","fetch_graph":"https://pith.science/api/pith-number/3KYWDOKOMWZFNSDNZHBLZYKTMZ/graph.json","fetch_events":"https://pith.science/api/pith-number/3KYWDOKOMWZFNSDNZHBLZYKTMZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3KYWDOKOMWZFNSDNZHBLZYKTMZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3KYWDOKOMWZFNSDNZHBLZYKTMZ/action/storage_attestation","attest_author":"https://pith.science/pith/3KYWDOKOMWZFNSDNZHBLZYKTMZ/action/author_attestation","sign_citation":"https://pith.science/pith/3KYWDOKOMWZFNSDNZHBLZYKTMZ/action/citation_signature","submit_replication":"https://pith.science/pith/3KYWDOKOMWZFNSDNZHBLZYKTMZ/action/replication_record"}},"created_at":"2026-07-05T08:55:08.112547+00:00","updated_at":"2026-07-05T08:55:08.112547+00:00"}