{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:ZRE5QRQZ5PNHMAAFZ5MGFP2PU2","short_pith_number":"pith:ZRE5QRQZ","schema_version":"1.0","canonical_sha256":"cc49d84619ebda760005cf5862bf4fa6bbbe02c9bb92fffef59d46b08cc7ca96","source":{"kind":"arxiv","id":"2606.07167","version":1},"attestation_state":"computed","paper":{"title":"UrduMMLU: A Massive Multitask Benchmark for Urdu Language Understanding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ahmer Tabassum, Hasan Iqbal, Momina Ahsan, Owais Aijaz, Preslav Nakov, Sarfraz Ahmad","submitted_at":"2026-06-05T11:35:27Z","abstract_excerpt":"Meaningful multilingual evaluation must test models in the target language and educational context. Urdu, spoken by more than 230 million people, lacks a broad MMLU-style benchmark built from native educational sources. We introduce UrduMMLU, a benchmark of 26,431 Urdu MCQs across 26 subjects and five domains, collected from native Urdu MCQ banks and public examination PDFs. Unlike translation-based resources, UrduMMLU covers both standard academic subjects and Urdu- and region-specific content. We label the exam-derived portion through dual human annotation with strict consensus filtering. We"},"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":"2606.07167","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-05T11:35:27Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"24b93b4eaec4d3d8ceb983ae4c9b31f2f2d9a5f6ea2cedc90590c65c6a65d797","abstract_canon_sha256":"67523c969f247832a83e2764758791c39706c4c18590b2a7cf03cfcaf62dc1c2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-08T01:04:50.394967Z","signature_b64":"JhhFEYjUU+FWLCzSHnTh9Hq42uSP7lDoDRiLfl9R/8N9DEj3Eq9JVElDzVFJeH3zsb1XOAIXWaE4bUFTApBAAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cc49d84619ebda760005cf5862bf4fa6bbbe02c9bb92fffef59d46b08cc7ca96","last_reissued_at":"2026-06-08T01:04:50.394037Z","signature_status":"signed_v1","first_computed_at":"2026-06-08T01:04:50.394037Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"UrduMMLU: A Massive Multitask Benchmark for Urdu Language Understanding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ahmer Tabassum, Hasan Iqbal, Momina Ahsan, Owais Aijaz, Preslav Nakov, Sarfraz Ahmad","submitted_at":"2026-06-05T11:35:27Z","abstract_excerpt":"Meaningful multilingual evaluation must test models in the target language and educational context. Urdu, spoken by more than 230 million people, lacks a broad MMLU-style benchmark built from native educational sources. We introduce UrduMMLU, a benchmark of 26,431 Urdu MCQs across 26 subjects and five domains, collected from native Urdu MCQ banks and public examination PDFs. Unlike translation-based resources, UrduMMLU covers both standard academic subjects and Urdu- and region-specific content. We label the exam-derived portion through dual human annotation with strict consensus filtering. We"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.07167","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/2606.07167/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":"2606.07167","created_at":"2026-06-08T01:04:50.394189+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.07167v1","created_at":"2026-06-08T01:04:50.394189+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.07167","created_at":"2026-06-08T01:04:50.394189+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZRE5QRQZ5PNH","created_at":"2026-06-08T01:04:50.394189+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZRE5QRQZ5PNHMAAF","created_at":"2026-06-08T01:04:50.394189+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZRE5QRQZ","created_at":"2026-06-08T01:04:50.394189+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ZRE5QRQZ5PNHMAAFZ5MGFP2PU2","json":"https://pith.science/pith/ZRE5QRQZ5PNHMAAFZ5MGFP2PU2.json","graph_json":"https://pith.science/api/pith-number/ZRE5QRQZ5PNHMAAFZ5MGFP2PU2/graph.json","events_json":"https://pith.science/api/pith-number/ZRE5QRQZ5PNHMAAFZ5MGFP2PU2/events.json","paper":"https://pith.science/paper/ZRE5QRQZ"},"agent_actions":{"view_html":"https://pith.science/pith/ZRE5QRQZ5PNHMAAFZ5MGFP2PU2","download_json":"https://pith.science/pith/ZRE5QRQZ5PNHMAAFZ5MGFP2PU2.json","view_paper":"https://pith.science/paper/ZRE5QRQZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.07167&json=true","fetch_graph":"https://pith.science/api/pith-number/ZRE5QRQZ5PNHMAAFZ5MGFP2PU2/graph.json","fetch_events":"https://pith.science/api/pith-number/ZRE5QRQZ5PNHMAAFZ5MGFP2PU2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZRE5QRQZ5PNHMAAFZ5MGFP2PU2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZRE5QRQZ5PNHMAAFZ5MGFP2PU2/action/storage_attestation","attest_author":"https://pith.science/pith/ZRE5QRQZ5PNHMAAFZ5MGFP2PU2/action/author_attestation","sign_citation":"https://pith.science/pith/ZRE5QRQZ5PNHMAAFZ5MGFP2PU2/action/citation_signature","submit_replication":"https://pith.science/pith/ZRE5QRQZ5PNHMAAFZ5MGFP2PU2/action/replication_record"}},"created_at":"2026-06-08T01:04:50.394189+00:00","updated_at":"2026-06-08T01:04:50.394189+00:00"}