{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:NUXNF7F3Z6JWPACQUBPLTUPOZ5","short_pith_number":"pith:NUXNF7F3","schema_version":"1.0","canonical_sha256":"6d2ed2fcbbcf93678050a05eb9d1eecf4b4b34ebe2b788ea351c934960039b79","source":{"kind":"arxiv","id":"2205.06356","version":2},"attestation_state":"computed","paper":{"title":"Beyond Static Models and Test Sets: Benchmarking the Potential of Pre-trained Models Across Tasks and Languages","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Kabir Ahuja, Monojit Choudhury, Sandipan Dandapat, Sunayana Sitaram","submitted_at":"2022-05-12T20:42:48Z","abstract_excerpt":"Although recent Massively Multilingual Language Models (MMLMs) like mBERT and XLMR support around 100 languages, most existing multilingual NLP benchmarks provide evaluation data in only a handful of these languages with little linguistic diversity. We argue that this makes the existing practices in multilingual evaluation unreliable and does not provide a full picture of the performance of MMLMs across the linguistic landscape. We propose that the recent work done in Performance Prediction for NLP tasks can serve as a potential solution in fixing benchmarking in Multilingual NLP by utilizing "},"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":"2205.06356","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-05-12T20:42:48Z","cross_cats_sorted":[],"title_canon_sha256":"c605b546d2f990b2fe2d883b6980f2c0b4ad2a22dd69af0cab79a687dd7bf683","abstract_canon_sha256":"b2730912d5ed2a970f26e8008c954957e6d91f3747fa9df49ed573e7631ca957"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:15:39.618552Z","signature_b64":"AAlAbx2ITJm7i38k299qWkGXg2EB3kQtIKX5FnyPxzjAWvCUspDfwD1thQBmXn2iKztjCSsOD2G13Ips6AszAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6d2ed2fcbbcf93678050a05eb9d1eecf4b4b34ebe2b788ea351c934960039b79","last_reissued_at":"2026-07-05T05:15:39.618091Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:15:39.618091Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Beyond Static Models and Test Sets: Benchmarking the Potential of Pre-trained Models Across Tasks and Languages","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Kabir Ahuja, Monojit Choudhury, Sandipan Dandapat, Sunayana Sitaram","submitted_at":"2022-05-12T20:42:48Z","abstract_excerpt":"Although recent Massively Multilingual Language Models (MMLMs) like mBERT and XLMR support around 100 languages, most existing multilingual NLP benchmarks provide evaluation data in only a handful of these languages with little linguistic diversity. We argue that this makes the existing practices in multilingual evaluation unreliable and does not provide a full picture of the performance of MMLMs across the linguistic landscape. We propose that the recent work done in Performance Prediction for NLP tasks can serve as a potential solution in fixing benchmarking in Multilingual NLP by utilizing "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2205.06356","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/2205.06356/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":"2205.06356","created_at":"2026-07-05T05:15:39.618153+00:00"},{"alias_kind":"arxiv_version","alias_value":"2205.06356v2","created_at":"2026-07-05T05:15:39.618153+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2205.06356","created_at":"2026-07-05T05:15:39.618153+00:00"},{"alias_kind":"pith_short_12","alias_value":"NUXNF7F3Z6JW","created_at":"2026-07-05T05:15:39.618153+00:00"},{"alias_kind":"pith_short_16","alias_value":"NUXNF7F3Z6JWPACQ","created_at":"2026-07-05T05:15:39.618153+00:00"},{"alias_kind":"pith_short_8","alias_value":"NUXNF7F3","created_at":"2026-07-05T05:15:39.618153+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2502.07101","citing_title":"SMAB: MAB based word Sensitivity Estimation Framework and its Applications in Adversarial Text Generation","ref_index":1,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NUXNF7F3Z6JWPACQUBPLTUPOZ5","json":"https://pith.science/pith/NUXNF7F3Z6JWPACQUBPLTUPOZ5.json","graph_json":"https://pith.science/api/pith-number/NUXNF7F3Z6JWPACQUBPLTUPOZ5/graph.json","events_json":"https://pith.science/api/pith-number/NUXNF7F3Z6JWPACQUBPLTUPOZ5/events.json","paper":"https://pith.science/paper/NUXNF7F3"},"agent_actions":{"view_html":"https://pith.science/pith/NUXNF7F3Z6JWPACQUBPLTUPOZ5","download_json":"https://pith.science/pith/NUXNF7F3Z6JWPACQUBPLTUPOZ5.json","view_paper":"https://pith.science/paper/NUXNF7F3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2205.06356&json=true","fetch_graph":"https://pith.science/api/pith-number/NUXNF7F3Z6JWPACQUBPLTUPOZ5/graph.json","fetch_events":"https://pith.science/api/pith-number/NUXNF7F3Z6JWPACQUBPLTUPOZ5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NUXNF7F3Z6JWPACQUBPLTUPOZ5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NUXNF7F3Z6JWPACQUBPLTUPOZ5/action/storage_attestation","attest_author":"https://pith.science/pith/NUXNF7F3Z6JWPACQUBPLTUPOZ5/action/author_attestation","sign_citation":"https://pith.science/pith/NUXNF7F3Z6JWPACQUBPLTUPOZ5/action/citation_signature","submit_replication":"https://pith.science/pith/NUXNF7F3Z6JWPACQUBPLTUPOZ5/action/replication_record"}},"created_at":"2026-07-05T05:15:39.618153+00:00","updated_at":"2026-07-05T05:15:39.618153+00:00"}