{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:N2EPT4AHNJLTNBT7JRE6XG3FK3","short_pith_number":"pith:N2EPT4AH","schema_version":"1.0","canonical_sha256":"6e88f9f0076a5736867f4c49eb9b6556c1a5714de1159c1aba58ebd8eae896a9","source":{"kind":"arxiv","id":"2503.18491","version":3},"attestation_state":"computed","paper":{"title":"MAGIC-VQA: Multimodal And Grounded Inference with Commonsense Knowledge for Visual Question Answering","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Eduard Hovy, Shuo Yang, Siwen Luo, Soyeon Caren Han","submitted_at":"2025-03-24T09:45:26Z","abstract_excerpt":"Visual Question Answering (VQA) requires reasoning across visual and textual modalities, yet Large Vision-Language Models (LVLMs) often lack integrated commonsense knowledge, limiting their robustness in real-world scenarios. To address this, we introduce MAGIC-VQA, a novel framework that enhances VQA by systematically integrating commonsense knowledge with LVLMs. MAGIC-VQA employs a three-stage process: (1) Explicit Knowledge Integration from external sources, (2) By-Type Post-Processing for contextual refinement, and (3) Implicit Knowledge Augmentation using a Graph Neural Network (GNN) for "},"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":"2503.18491","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-03-24T09:45:26Z","cross_cats_sorted":[],"title_canon_sha256":"6b0b03e7db1d1d7dc6823d7e19a9d3c17ac6af70b55eb4d6f936e739a687b1da","abstract_canon_sha256":"8f4f27da43ce764c0908f72a85f52f9f1e3e51c7e59ff2aeef47a4978512a78f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:19:26.049281Z","signature_b64":"jVKmi+oAZ+jvq1zm+4RGL3NB4KwAbC1D3wbONiaXCDAk0neX7CBy3g1LZS0JFlEx0nk9Jp7LwJuCO/u1EH6FAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6e88f9f0076a5736867f4c49eb9b6556c1a5714de1159c1aba58ebd8eae896a9","last_reissued_at":"2026-07-05T11:19:26.048790Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:19:26.048790Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MAGIC-VQA: Multimodal And Grounded Inference with Commonsense Knowledge for Visual Question Answering","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Eduard Hovy, Shuo Yang, Siwen Luo, Soyeon Caren Han","submitted_at":"2025-03-24T09:45:26Z","abstract_excerpt":"Visual Question Answering (VQA) requires reasoning across visual and textual modalities, yet Large Vision-Language Models (LVLMs) often lack integrated commonsense knowledge, limiting their robustness in real-world scenarios. To address this, we introduce MAGIC-VQA, a novel framework that enhances VQA by systematically integrating commonsense knowledge with LVLMs. MAGIC-VQA employs a three-stage process: (1) Explicit Knowledge Integration from external sources, (2) By-Type Post-Processing for contextual refinement, and (3) Implicit Knowledge Augmentation using a Graph Neural Network (GNN) for "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.18491","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/2503.18491/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":"2503.18491","created_at":"2026-07-05T11:19:26.048851+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.18491v3","created_at":"2026-07-05T11:19:26.048851+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.18491","created_at":"2026-07-05T11:19:26.048851+00:00"},{"alias_kind":"pith_short_12","alias_value":"N2EPT4AHNJLT","created_at":"2026-07-05T11:19:26.048851+00:00"},{"alias_kind":"pith_short_16","alias_value":"N2EPT4AHNJLTNBT7","created_at":"2026-07-05T11:19:26.048851+00:00"},{"alias_kind":"pith_short_8","alias_value":"N2EPT4AH","created_at":"2026-07-05T11:19:26.048851+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.17262","citing_title":"VisionTrap: Unanswerable Questions On Visual Data","ref_index":40,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/N2EPT4AHNJLTNBT7JRE6XG3FK3","json":"https://pith.science/pith/N2EPT4AHNJLTNBT7JRE6XG3FK3.json","graph_json":"https://pith.science/api/pith-number/N2EPT4AHNJLTNBT7JRE6XG3FK3/graph.json","events_json":"https://pith.science/api/pith-number/N2EPT4AHNJLTNBT7JRE6XG3FK3/events.json","paper":"https://pith.science/paper/N2EPT4AH"},"agent_actions":{"view_html":"https://pith.science/pith/N2EPT4AHNJLTNBT7JRE6XG3FK3","download_json":"https://pith.science/pith/N2EPT4AHNJLTNBT7JRE6XG3FK3.json","view_paper":"https://pith.science/paper/N2EPT4AH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.18491&json=true","fetch_graph":"https://pith.science/api/pith-number/N2EPT4AHNJLTNBT7JRE6XG3FK3/graph.json","fetch_events":"https://pith.science/api/pith-number/N2EPT4AHNJLTNBT7JRE6XG3FK3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/N2EPT4AHNJLTNBT7JRE6XG3FK3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/N2EPT4AHNJLTNBT7JRE6XG3FK3/action/storage_attestation","attest_author":"https://pith.science/pith/N2EPT4AHNJLTNBT7JRE6XG3FK3/action/author_attestation","sign_citation":"https://pith.science/pith/N2EPT4AHNJLTNBT7JRE6XG3FK3/action/citation_signature","submit_replication":"https://pith.science/pith/N2EPT4AHNJLTNBT7JRE6XG3FK3/action/replication_record"}},"created_at":"2026-07-05T11:19:26.048851+00:00","updated_at":"2026-07-05T11:19:26.048851+00:00"}