{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:THFWHQZ5JXA23CAXUXRLKUJB3F","short_pith_number":"pith:THFWHQZ5","schema_version":"1.0","canonical_sha256":"99cb63c33d4dc1ad8817a5e2b55121d966cf3bdaad1e09a4aa899f1d94381b0f","source":{"kind":"arxiv","id":"2404.13591","version":2},"attestation_state":"computed","paper":{"title":"MARVEL: Multidimensional Abstraction and Reasoning through Visual Evaluation and Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Filip Ilievski, Jay Pujara, Jiarui Zhang, Kaixin Ma, Kexuan Sun, Kian Ahrabian, Yifan Jiang, Zhivar Sourati","submitted_at":"2024-04-21T09:15:02Z","abstract_excerpt":"While multi-modal large language models (MLLMs) have shown significant progress on many popular visual reasoning benchmarks, whether they possess abstract visual reasoning abilities remains an open question. Similar to the Sudoku puzzles, abstract visual reasoning (AVR) problems require finding high-level patterns (e.g., repetition constraints) that control the input shapes (e.g., digits) in a specific task configuration (e.g., matrix). However, existing AVR benchmarks only considered a limited set of patterns (addition, conjunction), input shapes (rectangle, square), and task configurations ("},"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":"2404.13591","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-04-21T09:15:02Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"e7d0abba2e49b2c3db0546a392c17290ba6a10c5d583fcb4c7276ffdfe14ec5e","abstract_canon_sha256":"3773e07d40c62695b6d3991f71afb386ba8029a2b388c70b241a45c5a4e0cfe0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:11:59.566838Z","signature_b64":"ne4ume7JwhwpxPIoTS33w9jxTRW+baXgyMVNuh/0ExJYtCNPQrikIP1/F1hNnCoQnsS1Xnkx/5VCZBGZC1GqBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"99cb63c33d4dc1ad8817a5e2b55121d966cf3bdaad1e09a4aa899f1d94381b0f","last_reissued_at":"2026-07-05T08:11:59.566340Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:11:59.566340Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MARVEL: Multidimensional Abstraction and Reasoning through Visual Evaluation and Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Filip Ilievski, Jay Pujara, Jiarui Zhang, Kaixin Ma, Kexuan Sun, Kian Ahrabian, Yifan Jiang, Zhivar Sourati","submitted_at":"2024-04-21T09:15:02Z","abstract_excerpt":"While multi-modal large language models (MLLMs) have shown significant progress on many popular visual reasoning benchmarks, whether they possess abstract visual reasoning abilities remains an open question. Similar to the Sudoku puzzles, abstract visual reasoning (AVR) problems require finding high-level patterns (e.g., repetition constraints) that control the input shapes (e.g., digits) in a specific task configuration (e.g., matrix). However, existing AVR benchmarks only considered a limited set of patterns (addition, conjunction), input shapes (rectangle, square), and task configurations ("},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.13591","kind":"arxiv","version":2},"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/2404.13591/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":"2404.13591","created_at":"2026-07-05T08:11:59.566401+00:00"},{"alias_kind":"arxiv_version","alias_value":"2404.13591v2","created_at":"2026-07-05T08:11:59.566401+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.13591","created_at":"2026-07-05T08:11:59.566401+00:00"},{"alias_kind":"pith_short_12","alias_value":"THFWHQZ5JXA2","created_at":"2026-07-05T08:11:59.566401+00:00"},{"alias_kind":"pith_short_16","alias_value":"THFWHQZ5JXA23CAX","created_at":"2026-07-05T08:11:59.566401+00:00"},{"alias_kind":"pith_short_8","alias_value":"THFWHQZ5","created_at":"2026-07-05T08:11:59.566401+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.21850","citing_title":"Beyond Perception: Evaluating Abstract Visual Reasoning through Multi-Stage Task","ref_index":18,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/THFWHQZ5JXA23CAXUXRLKUJB3F","json":"https://pith.science/pith/THFWHQZ5JXA23CAXUXRLKUJB3F.json","graph_json":"https://pith.science/api/pith-number/THFWHQZ5JXA23CAXUXRLKUJB3F/graph.json","events_json":"https://pith.science/api/pith-number/THFWHQZ5JXA23CAXUXRLKUJB3F/events.json","paper":"https://pith.science/paper/THFWHQZ5"},"agent_actions":{"view_html":"https://pith.science/pith/THFWHQZ5JXA23CAXUXRLKUJB3F","download_json":"https://pith.science/pith/THFWHQZ5JXA23CAXUXRLKUJB3F.json","view_paper":"https://pith.science/paper/THFWHQZ5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2404.13591&json=true","fetch_graph":"https://pith.science/api/pith-number/THFWHQZ5JXA23CAXUXRLKUJB3F/graph.json","fetch_events":"https://pith.science/api/pith-number/THFWHQZ5JXA23CAXUXRLKUJB3F/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/THFWHQZ5JXA23CAXUXRLKUJB3F/action/timestamp_anchor","attest_storage":"https://pith.science/pith/THFWHQZ5JXA23CAXUXRLKUJB3F/action/storage_attestation","attest_author":"https://pith.science/pith/THFWHQZ5JXA23CAXUXRLKUJB3F/action/author_attestation","sign_citation":"https://pith.science/pith/THFWHQZ5JXA23CAXUXRLKUJB3F/action/citation_signature","submit_replication":"https://pith.science/pith/THFWHQZ5JXA23CAXUXRLKUJB3F/action/replication_record"}},"created_at":"2026-07-05T08:11:59.566401+00:00","updated_at":"2026-07-05T08:11:59.566401+00:00"}