{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:BZAJDH2GCB6FLCULNRY7QRWEEG","short_pith_number":"pith:BZAJDH2G","schema_version":"1.0","canonical_sha256":"0e40919f46107c558a8b6c71f846c421b7ea4e2fae57142c568475a0ecab66ee","source":{"kind":"arxiv","id":"2408.04810","version":1},"attestation_state":"computed","paper":{"title":"UniBench: Visual Reasoning Requires Rethinking Vision-Language Beyond Scaling","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Caner Hazirbas, Diane Bouchacourt, Haider Al-Tahan, Mark Ibrahim, Quentin Garrido, Randall Balestriero","submitted_at":"2024-08-09T01:41:05Z","abstract_excerpt":"Significant research efforts have been made to scale and improve vision-language model (VLM) training approaches. Yet, with an ever-growing number of benchmarks, researchers are tasked with the heavy burden of implementing each protocol, bearing a non-trivial computational cost, and making sense of how all these benchmarks translate into meaningful axes of progress. To facilitate a systematic evaluation of VLM progress, we introduce UniBench: a unified implementation of 50+ VLM benchmarks spanning a comprehensive range of carefully categorized capabilities from object recognition to spatial aw"},"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":"2408.04810","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-08-09T01:41:05Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"6a237ff5526304ca3bcaf47dd72006b0cd11c409694e4be2de0a0e995df6ad55","abstract_canon_sha256":"972384e77f9ca585bb693e1430b4e196aa166e1ff77b3fcd5cccbc643e751341"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:53:44.397712Z","signature_b64":"hJrSzC+r9pEbn0jAYdAZBZFVa2Yrc+HYwu9M1/IBJF/+JpjG92Ci5KgkaCBxUECzSWg/l8IzvqTr3SgvN99/Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0e40919f46107c558a8b6c71f846c421b7ea4e2fae57142c568475a0ecab66ee","last_reissued_at":"2026-07-05T08:53:44.397224Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:53:44.397224Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"UniBench: Visual Reasoning Requires Rethinking Vision-Language Beyond Scaling","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Caner Hazirbas, Diane Bouchacourt, Haider Al-Tahan, Mark Ibrahim, Quentin Garrido, Randall Balestriero","submitted_at":"2024-08-09T01:41:05Z","abstract_excerpt":"Significant research efforts have been made to scale and improve vision-language model (VLM) training approaches. Yet, with an ever-growing number of benchmarks, researchers are tasked with the heavy burden of implementing each protocol, bearing a non-trivial computational cost, and making sense of how all these benchmarks translate into meaningful axes of progress. To facilitate a systematic evaluation of VLM progress, we introduce UniBench: a unified implementation of 50+ VLM benchmarks spanning a comprehensive range of carefully categorized capabilities from object recognition to spatial aw"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.04810","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/2408.04810/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":"2408.04810","created_at":"2026-07-05T08:53:44.397281+00:00"},{"alias_kind":"arxiv_version","alias_value":"2408.04810v1","created_at":"2026-07-05T08:53:44.397281+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.04810","created_at":"2026-07-05T08:53:44.397281+00:00"},{"alias_kind":"pith_short_12","alias_value":"BZAJDH2GCB6F","created_at":"2026-07-05T08:53:44.397281+00:00"},{"alias_kind":"pith_short_16","alias_value":"BZAJDH2GCB6FLCUL","created_at":"2026-07-05T08:53:44.397281+00:00"},{"alias_kind":"pith_short_8","alias_value":"BZAJDH2G","created_at":"2026-07-05T08:53:44.397281+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2507.01955","citing_title":"How Well Does GPT-4o Understand Vision? Evaluating Multimodal Foundation Models on Standard Computer Vision Tasks","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2604.03893","citing_title":"FeynmanBench: Benchmarking Multimodal LLMs on Diagrammatic Physics Reasoning","ref_index":1,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BZAJDH2GCB6FLCULNRY7QRWEEG","json":"https://pith.science/pith/BZAJDH2GCB6FLCULNRY7QRWEEG.json","graph_json":"https://pith.science/api/pith-number/BZAJDH2GCB6FLCULNRY7QRWEEG/graph.json","events_json":"https://pith.science/api/pith-number/BZAJDH2GCB6FLCULNRY7QRWEEG/events.json","paper":"https://pith.science/paper/BZAJDH2G"},"agent_actions":{"view_html":"https://pith.science/pith/BZAJDH2GCB6FLCULNRY7QRWEEG","download_json":"https://pith.science/pith/BZAJDH2GCB6FLCULNRY7QRWEEG.json","view_paper":"https://pith.science/paper/BZAJDH2G","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2408.04810&json=true","fetch_graph":"https://pith.science/api/pith-number/BZAJDH2GCB6FLCULNRY7QRWEEG/graph.json","fetch_events":"https://pith.science/api/pith-number/BZAJDH2GCB6FLCULNRY7QRWEEG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BZAJDH2GCB6FLCULNRY7QRWEEG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BZAJDH2GCB6FLCULNRY7QRWEEG/action/storage_attestation","attest_author":"https://pith.science/pith/BZAJDH2GCB6FLCULNRY7QRWEEG/action/author_attestation","sign_citation":"https://pith.science/pith/BZAJDH2GCB6FLCULNRY7QRWEEG/action/citation_signature","submit_replication":"https://pith.science/pith/BZAJDH2GCB6FLCULNRY7QRWEEG/action/replication_record"}},"created_at":"2026-07-05T08:53:44.397281+00:00","updated_at":"2026-07-05T08:53:44.397281+00:00"}