{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:3ODL5KLAKGGNBLNUIAIWZQAASR","short_pith_number":"pith:3ODL5KLA","schema_version":"1.0","canonical_sha256":"db86bea960518cd0adb440116cc00094711f82754361335857b7c59a4ea3a1a8","source":{"kind":"arxiv","id":"2410.03568","version":1},"attestation_state":"computed","paper":{"title":"Towards Linguistically-Aware and Language-Independent Tokenization for Large Language Models (LLMs)","license":"http://creativecommons.org/publicdomain/zero/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Abrar Rahman, Binit Mohanty, Garry Bowlin, Sean McGunigal","submitted_at":"2024-10-04T16:18:29Z","abstract_excerpt":"This paper presents a comprehensive study on the tokenization techniques employed by state-of-the-art large language models (LLMs) and their implications on the cost and availability of services across different languages, especially low resource languages. The analysis considers multiple LLMs, including GPT-4 (using cl100k_base embeddings), GPT-3 (with p50k_base embeddings), and DaVinci (employing r50k_base embeddings), as well as the widely used BERT base tokenizer. The study evaluates the tokenization variability observed across these models and investigates the challenges of linguistic rep"},"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":"2410.03568","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/publicdomain/zero/1.0/","primary_cat":"cs.CL","submitted_at":"2024-10-04T16:18:29Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"3094e47dcf0924b36a1cb546b3dd505e72db3352c00bfa9a24727c90e301790f","abstract_canon_sha256":"3a8dc79122daac1b49246a518cc4f8194a9952bda29c18c1f7e37ab3b9423113"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:16:03.078691Z","signature_b64":"ZibIY77GV5ceCVNFA21z3EUrZmThX3T8UgqrPqVLEXkA0Of3/72SS0WVQUD/K/YoCQjZ56YAfnLymdABR4ByAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"db86bea960518cd0adb440116cc00094711f82754361335857b7c59a4ea3a1a8","last_reissued_at":"2026-07-05T09:16:03.078251Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:16:03.078251Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Towards Linguistically-Aware and Language-Independent Tokenization for Large Language Models (LLMs)","license":"http://creativecommons.org/publicdomain/zero/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Abrar Rahman, Binit Mohanty, Garry Bowlin, Sean McGunigal","submitted_at":"2024-10-04T16:18:29Z","abstract_excerpt":"This paper presents a comprehensive study on the tokenization techniques employed by state-of-the-art large language models (LLMs) and their implications on the cost and availability of services across different languages, especially low resource languages. The analysis considers multiple LLMs, including GPT-4 (using cl100k_base embeddings), GPT-3 (with p50k_base embeddings), and DaVinci (employing r50k_base embeddings), as well as the widely used BERT base tokenizer. The study evaluates the tokenization variability observed across these models and investigates the challenges of linguistic rep"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.03568","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/2410.03568/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":"2410.03568","created_at":"2026-07-05T09:16:03.078305+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.03568v1","created_at":"2026-07-05T09:16:03.078305+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.03568","created_at":"2026-07-05T09:16:03.078305+00:00"},{"alias_kind":"pith_short_12","alias_value":"3ODL5KLAKGGN","created_at":"2026-07-05T09:16:03.078305+00:00"},{"alias_kind":"pith_short_16","alias_value":"3ODL5KLAKGGNBLNU","created_at":"2026-07-05T09:16:03.078305+00:00"},{"alias_kind":"pith_short_8","alias_value":"3ODL5KLA","created_at":"2026-07-05T09:16:03.078305+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2412.18626","citing_title":"Why Do Large Language Models (LLMs) Struggle to Count Letters?","ref_index":14,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3ODL5KLAKGGNBLNUIAIWZQAASR","json":"https://pith.science/pith/3ODL5KLAKGGNBLNUIAIWZQAASR.json","graph_json":"https://pith.science/api/pith-number/3ODL5KLAKGGNBLNUIAIWZQAASR/graph.json","events_json":"https://pith.science/api/pith-number/3ODL5KLAKGGNBLNUIAIWZQAASR/events.json","paper":"https://pith.science/paper/3ODL5KLA"},"agent_actions":{"view_html":"https://pith.science/pith/3ODL5KLAKGGNBLNUIAIWZQAASR","download_json":"https://pith.science/pith/3ODL5KLAKGGNBLNUIAIWZQAASR.json","view_paper":"https://pith.science/paper/3ODL5KLA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.03568&json=true","fetch_graph":"https://pith.science/api/pith-number/3ODL5KLAKGGNBLNUIAIWZQAASR/graph.json","fetch_events":"https://pith.science/api/pith-number/3ODL5KLAKGGNBLNUIAIWZQAASR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3ODL5KLAKGGNBLNUIAIWZQAASR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3ODL5KLAKGGNBLNUIAIWZQAASR/action/storage_attestation","attest_author":"https://pith.science/pith/3ODL5KLAKGGNBLNUIAIWZQAASR/action/author_attestation","sign_citation":"https://pith.science/pith/3ODL5KLAKGGNBLNUIAIWZQAASR/action/citation_signature","submit_replication":"https://pith.science/pith/3ODL5KLAKGGNBLNUIAIWZQAASR/action/replication_record"}},"created_at":"2026-07-05T09:16:03.078305+00:00","updated_at":"2026-07-05T09:16:03.078305+00:00"}