{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:Q654GWMPBWZLQ6MAYAVRDZNFUM","short_pith_number":"pith:Q654GWMP","schema_version":"1.0","canonical_sha256":"87bbc3598f0db2b87980c02b11e5a5a311cc560faeb4e7def77451d4df9592b1","source":{"kind":"arxiv","id":"2408.10441","version":3},"attestation_state":"computed","paper":{"title":"Goldfish: Monolingual Language Models for 350 Languages","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Benjamin K. Bergen, Catherine Arnett, Tyler A. Chang, Zhuowen Tu","submitted_at":"2024-08-19T22:31:21Z","abstract_excerpt":"For many low-resource languages, the only available language models are large multilingual models trained on many languages simultaneously. Despite state-of-the-art performance on reasoning tasks, we find that these models still struggle with basic grammatical text generation in many languages. First, large multilingual models perform worse than bigrams for many languages (e.g. 24% of languages in XGLM 4.5B; 43% in BLOOM 7.1B) using FLORES perplexity as an evaluation metric. Second, when we train small monolingual models with only 125M parameters on 1GB or less data for 350 languages, these sm"},"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":"2408.10441","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-08-19T22:31:21Z","cross_cats_sorted":[],"title_canon_sha256":"cfaff8442ceb980a006fc50fc93a4b708612aae88a7708efe032f491ebabf72b","abstract_canon_sha256":"978eaa2eaa28fde83ddabc5c9a4f61f4651f5f53d77ebd501ac33f7787f1d484"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-01T01:02:12.808563Z","signature_b64":"yS9dTN3W5Kso5VKnd4N1Evd8QyaBrob9ommjmtJ01srCnYcCqqOdNccPLimIh8culUxlxcGh6Mhxa7w7m05eBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"87bbc3598f0db2b87980c02b11e5a5a311cc560faeb4e7def77451d4df9592b1","last_reissued_at":"2026-06-01T01:02:12.807206Z","signature_status":"signed_v1","first_computed_at":"2026-06-01T01:02:12.807206Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Goldfish: Monolingual Language Models for 350 Languages","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Benjamin K. Bergen, Catherine Arnett, Tyler A. Chang, Zhuowen Tu","submitted_at":"2024-08-19T22:31:21Z","abstract_excerpt":"For many low-resource languages, the only available language models are large multilingual models trained on many languages simultaneously. Despite state-of-the-art performance on reasoning tasks, we find that these models still struggle with basic grammatical text generation in many languages. First, large multilingual models perform worse than bigrams for many languages (e.g. 24% of languages in XGLM 4.5B; 43% in BLOOM 7.1B) using FLORES perplexity as an evaluation metric. Second, when we train small monolingual models with only 125M parameters on 1GB or less data for 350 languages, these sm"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.10441","kind":"arxiv","version":3},"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/2408.10441/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":"2408.10441","created_at":"2026-06-01T01:02:12.807377+00:00"},{"alias_kind":"arxiv_version","alias_value":"2408.10441v3","created_at":"2026-06-01T01:02:12.807377+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.10441","created_at":"2026-06-01T01:02:12.807377+00:00"},{"alias_kind":"pith_short_12","alias_value":"Q654GWMPBWZL","created_at":"2026-06-01T01:02:12.807377+00:00"},{"alias_kind":"pith_short_16","alias_value":"Q654GWMPBWZLQ6MA","created_at":"2026-06-01T01:02:12.807377+00:00"},{"alias_kind":"pith_short_8","alias_value":"Q654GWMP","created_at":"2026-06-01T01:02:12.807377+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.02132","citing_title":"Model Internal Sleuthing: Finding Lexical Identity and Inflectional Features in Modern Language Models","ref_index":7,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/Q654GWMPBWZLQ6MAYAVRDZNFUM","json":"https://pith.science/pith/Q654GWMPBWZLQ6MAYAVRDZNFUM.json","graph_json":"https://pith.science/api/pith-number/Q654GWMPBWZLQ6MAYAVRDZNFUM/graph.json","events_json":"https://pith.science/api/pith-number/Q654GWMPBWZLQ6MAYAVRDZNFUM/events.json","paper":"https://pith.science/paper/Q654GWMP"},"agent_actions":{"view_html":"https://pith.science/pith/Q654GWMPBWZLQ6MAYAVRDZNFUM","download_json":"https://pith.science/pith/Q654GWMPBWZLQ6MAYAVRDZNFUM.json","view_paper":"https://pith.science/paper/Q654GWMP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2408.10441&json=true","fetch_graph":"https://pith.science/api/pith-number/Q654GWMPBWZLQ6MAYAVRDZNFUM/graph.json","fetch_events":"https://pith.science/api/pith-number/Q654GWMPBWZLQ6MAYAVRDZNFUM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/Q654GWMPBWZLQ6MAYAVRDZNFUM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/Q654GWMPBWZLQ6MAYAVRDZNFUM/action/storage_attestation","attest_author":"https://pith.science/pith/Q654GWMPBWZLQ6MAYAVRDZNFUM/action/author_attestation","sign_citation":"https://pith.science/pith/Q654GWMPBWZLQ6MAYAVRDZNFUM/action/citation_signature","submit_replication":"https://pith.science/pith/Q654GWMPBWZLQ6MAYAVRDZNFUM/action/replication_record"}},"created_at":"2026-06-01T01:02:12.807377+00:00","updated_at":"2026-06-01T01:02:12.807377+00:00"}