{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:2WINJSKGPGORRGU46S7F6PLQ67","short_pith_number":"pith:2WINJSKG","schema_version":"1.0","canonical_sha256":"d590d4c946799d189a9cf4be5f3d70f7e135a9e70357d67e2550ada4baec9be9","source":{"kind":"arxiv","id":"2508.21294","version":1},"attestation_state":"computed","paper":{"title":"BLUEX Revisited: Enhancing Benchmark Coverage with Automatic Captioning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Giovana Kerche Bon\\'as, Jo\\~ao Guilherme Alves Santos, Thales Sales Almeida","submitted_at":"2025-08-29T01:23:28Z","abstract_excerpt":"With the growing capabilities of Large Language Models (LLMs), there is an increasing need for robust evaluation methods, especially in multilingual and non-English contexts. We present an updated version of the BLUEX dataset, now including 2024-2025 exams and automatically generated image captions using state-of-the-art models, enhancing its relevance for data contamination studies in LLM pretraining. Captioning strategies increase accessibility to text-only models by more than 40%, producing 1,422 usable questions, more than doubling the number in the original BLUEX. We evaluated commercial "},"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":"2508.21294","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-08-29T01:23:28Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"c796216139561dd66d8a0633d39daaf2ecab2191ca754c168fc24f4661b4e166","abstract_canon_sha256":"551fa50d82c75b6e004852358c88729fbbe5ad520219145868152862f845f79c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:01:32.086030Z","signature_b64":"X59/Fv3PXUKcQWrFLad1f6/4rEPvJKOfLlsQr+LbO4A5UeXhdd/4gIzUBd3GN+e0ergjXur4Z2DBK0rVH7VBCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d590d4c946799d189a9cf4be5f3d70f7e135a9e70357d67e2550ada4baec9be9","last_reissued_at":"2026-07-05T12:01:32.085513Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:01:32.085513Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"BLUEX Revisited: Enhancing Benchmark Coverage with Automatic Captioning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Giovana Kerche Bon\\'as, Jo\\~ao Guilherme Alves Santos, Thales Sales Almeida","submitted_at":"2025-08-29T01:23:28Z","abstract_excerpt":"With the growing capabilities of Large Language Models (LLMs), there is an increasing need for robust evaluation methods, especially in multilingual and non-English contexts. We present an updated version of the BLUEX dataset, now including 2024-2025 exams and automatically generated image captions using state-of-the-art models, enhancing its relevance for data contamination studies in LLM pretraining. Captioning strategies increase accessibility to text-only models by more than 40%, producing 1,422 usable questions, more than doubling the number in the original BLUEX. We evaluated commercial "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.21294","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/2508.21294/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":"2508.21294","created_at":"2026-07-05T12:01:32.085572+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.21294v1","created_at":"2026-07-05T12:01:32.085572+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.21294","created_at":"2026-07-05T12:01:32.085572+00:00"},{"alias_kind":"pith_short_12","alias_value":"2WINJSKGPGOR","created_at":"2026-07-05T12:01:32.085572+00:00"},{"alias_kind":"pith_short_16","alias_value":"2WINJSKGPGORRGU4","created_at":"2026-07-05T12:01:32.085572+00:00"},{"alias_kind":"pith_short_8","alias_value":"2WINJSKG","created_at":"2026-07-05T12:01:32.085572+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.14448","citing_title":"MARCA: A Checklist-Based Benchmark for Multilingual Web Search","ref_index":24,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2WINJSKGPGORRGU46S7F6PLQ67","json":"https://pith.science/pith/2WINJSKGPGORRGU46S7F6PLQ67.json","graph_json":"https://pith.science/api/pith-number/2WINJSKGPGORRGU46S7F6PLQ67/graph.json","events_json":"https://pith.science/api/pith-number/2WINJSKGPGORRGU46S7F6PLQ67/events.json","paper":"https://pith.science/paper/2WINJSKG"},"agent_actions":{"view_html":"https://pith.science/pith/2WINJSKGPGORRGU46S7F6PLQ67","download_json":"https://pith.science/pith/2WINJSKGPGORRGU46S7F6PLQ67.json","view_paper":"https://pith.science/paper/2WINJSKG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.21294&json=true","fetch_graph":"https://pith.science/api/pith-number/2WINJSKGPGORRGU46S7F6PLQ67/graph.json","fetch_events":"https://pith.science/api/pith-number/2WINJSKGPGORRGU46S7F6PLQ67/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2WINJSKGPGORRGU46S7F6PLQ67/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2WINJSKGPGORRGU46S7F6PLQ67/action/storage_attestation","attest_author":"https://pith.science/pith/2WINJSKGPGORRGU46S7F6PLQ67/action/author_attestation","sign_citation":"https://pith.science/pith/2WINJSKGPGORRGU46S7F6PLQ67/action/citation_signature","submit_replication":"https://pith.science/pith/2WINJSKGPGORRGU46S7F6PLQ67/action/replication_record"}},"created_at":"2026-07-05T12:01:32.085572+00:00","updated_at":"2026-07-05T12:01:32.085572+00:00"}