{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:NJ4S4D5TK27ZWSBQRAKICKZNEB","short_pith_number":"pith:NJ4S4D5T","schema_version":"1.0","canonical_sha256":"6a792e0fb356bf9b48308814812b2d20752696a976d3ee7d45bb25599369a3ba","source":{"kind":"arxiv","id":"2411.14198","version":1},"attestation_state":"computed","paper":{"title":"Why do language models perform worse for morphologically complex languages?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Benjamin K. Bergen, Catherine Arnett","submitted_at":"2024-11-21T15:06:51Z","abstract_excerpt":"Language models perform differently across languages. It has been previously suggested that morphological typology may explain some of this variability (Cotterell et al., 2018). We replicate previous analyses and find additional new evidence for a performance gap between agglutinative and fusional languages, where fusional languages, such as English, tend to have better language modeling performance than morphologically more complex languages like Turkish. We then propose and test three possible causes for this performance gap: morphological alignment of tokenizers, tokenization quality, and d"},"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":"2411.14198","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-11-21T15:06:51Z","cross_cats_sorted":[],"title_canon_sha256":"c6e757d8d1424fbaf31d1dbf93b1b1c49e5b54fe2d80be2dd08845131ab1bb2a","abstract_canon_sha256":"9dd095ee4445467ff69286c0d368500bd61b26bb8b60f6a7e94f321f71e1b9b5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:38:43.074743Z","signature_b64":"hiQ7BrfX+d+ZKMUNjhlgHNkOGZugaMWbnygmIvlmro3jQ+npoO+vVZSGrQPsXqDXtO9v6sQKU2YHjbD9HorGCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6a792e0fb356bf9b48308814812b2d20752696a976d3ee7d45bb25599369a3ba","last_reissued_at":"2026-07-05T09:38:43.074251Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:38:43.074251Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Why do language models perform worse for morphologically complex languages?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Benjamin K. Bergen, Catherine Arnett","submitted_at":"2024-11-21T15:06:51Z","abstract_excerpt":"Language models perform differently across languages. It has been previously suggested that morphological typology may explain some of this variability (Cotterell et al., 2018). We replicate previous analyses and find additional new evidence for a performance gap between agglutinative and fusional languages, where fusional languages, such as English, tend to have better language modeling performance than morphologically more complex languages like Turkish. We then propose and test three possible causes for this performance gap: morphological alignment of tokenizers, tokenization quality, and d"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.14198","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/2411.14198/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":"2411.14198","created_at":"2026-07-05T09:38:43.074309+00:00"},{"alias_kind":"arxiv_version","alias_value":"2411.14198v1","created_at":"2026-07-05T09:38:43.074309+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.14198","created_at":"2026-07-05T09:38:43.074309+00:00"},{"alias_kind":"pith_short_12","alias_value":"NJ4S4D5TK27Z","created_at":"2026-07-05T09:38:43.074309+00:00"},{"alias_kind":"pith_short_16","alias_value":"NJ4S4D5TK27ZWSBQ","created_at":"2026-07-05T09:38:43.074309+00:00"},{"alias_kind":"pith_short_8","alias_value":"NJ4S4D5T","created_at":"2026-07-05T09:38:43.074309+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.22582","citing_title":"Less, but Better: Efficient Multilingual Expansion for LLMs via Layer-wise Mixture-of-Experts","ref_index":4,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NJ4S4D5TK27ZWSBQRAKICKZNEB","json":"https://pith.science/pith/NJ4S4D5TK27ZWSBQRAKICKZNEB.json","graph_json":"https://pith.science/api/pith-number/NJ4S4D5TK27ZWSBQRAKICKZNEB/graph.json","events_json":"https://pith.science/api/pith-number/NJ4S4D5TK27ZWSBQRAKICKZNEB/events.json","paper":"https://pith.science/paper/NJ4S4D5T"},"agent_actions":{"view_html":"https://pith.science/pith/NJ4S4D5TK27ZWSBQRAKICKZNEB","download_json":"https://pith.science/pith/NJ4S4D5TK27ZWSBQRAKICKZNEB.json","view_paper":"https://pith.science/paper/NJ4S4D5T","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2411.14198&json=true","fetch_graph":"https://pith.science/api/pith-number/NJ4S4D5TK27ZWSBQRAKICKZNEB/graph.json","fetch_events":"https://pith.science/api/pith-number/NJ4S4D5TK27ZWSBQRAKICKZNEB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NJ4S4D5TK27ZWSBQRAKICKZNEB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NJ4S4D5TK27ZWSBQRAKICKZNEB/action/storage_attestation","attest_author":"https://pith.science/pith/NJ4S4D5TK27ZWSBQRAKICKZNEB/action/author_attestation","sign_citation":"https://pith.science/pith/NJ4S4D5TK27ZWSBQRAKICKZNEB/action/citation_signature","submit_replication":"https://pith.science/pith/NJ4S4D5TK27ZWSBQRAKICKZNEB/action/replication_record"}},"created_at":"2026-07-05T09:38:43.074309+00:00","updated_at":"2026-07-05T09:38:43.074309+00:00"}