{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:345KRE4MNBEDTW6ZDKYYBK6O7N","short_pith_number":"pith:345KRE4M","schema_version":"1.0","canonical_sha256":"df3aa8938c684839dbd91ab180abcefb6c3ff725369678bd957627756d33e682","source":{"kind":"arxiv","id":"2310.11628","version":1},"attestation_state":"computed","paper":{"title":"Learn Your Tokens: Word-Pooled Tokenization for Language Modeling","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Avijit Thawani, Jay Pujara, Saurabh Ghanekar, Xiaoyuan Zhu","submitted_at":"2023-10-17T23:34:39Z","abstract_excerpt":"Language models typically tokenize text into subwords, using a deterministic, hand-engineered heuristic of combining characters into longer surface-level strings such as 'ing' or whole words. Recent literature has repeatedly shown the limitations of such a tokenization strategy, particularly for documents not written in English and for representing numbers. On the other extreme, byte/character-level language models are much less restricted but suffer from increased sequence description lengths and a subsequent quadratic expansion in self-attention computation. Recent attempts to compress and l"},"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":"2310.11628","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-10-17T23:34:39Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"b1ce45cd20c6fd2cf159de2803f03861864d4bf64c3e6a077f304f2a2ec54f0c","abstract_canon_sha256":"287f98af7c165319c9df1b46d3963b2d55a079e759764579bf392f617008cae2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:02:10.254991Z","signature_b64":"XYgIlYZzKzerSlhlfNZJ59OrTBGFpak5H+IVIH/2ebxmaQQedNG5E85I5QAfrqPu3J+j4eetvw4R1Gyl5RAlDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"df3aa8938c684839dbd91ab180abcefb6c3ff725369678bd957627756d33e682","last_reissued_at":"2026-07-05T07:02:10.254584Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:02:10.254584Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learn Your Tokens: Word-Pooled Tokenization for Language Modeling","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Avijit Thawani, Jay Pujara, Saurabh Ghanekar, Xiaoyuan Zhu","submitted_at":"2023-10-17T23:34:39Z","abstract_excerpt":"Language models typically tokenize text into subwords, using a deterministic, hand-engineered heuristic of combining characters into longer surface-level strings such as 'ing' or whole words. Recent literature has repeatedly shown the limitations of such a tokenization strategy, particularly for documents not written in English and for representing numbers. On the other extreme, byte/character-level language models are much less restricted but suffer from increased sequence description lengths and a subsequent quadratic expansion in self-attention computation. Recent attempts to compress and l"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.11628","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/2310.11628/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":"2310.11628","created_at":"2026-07-05T07:02:10.254641+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.11628v1","created_at":"2026-07-05T07:02:10.254641+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.11628","created_at":"2026-07-05T07:02:10.254641+00:00"},{"alias_kind":"pith_short_12","alias_value":"345KRE4MNBED","created_at":"2026-07-05T07:02:10.254641+00:00"},{"alias_kind":"pith_short_16","alias_value":"345KRE4MNBEDTW6Z","created_at":"2026-07-05T07:02:10.254641+00:00"},{"alias_kind":"pith_short_8","alias_value":"345KRE4M","created_at":"2026-07-05T07:02:10.254641+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2411.17679","citing_title":"Enhancing Character-Level Understanding in LLMs through Token Internal Structure Learning","ref_index":29,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/345KRE4MNBEDTW6ZDKYYBK6O7N","json":"https://pith.science/pith/345KRE4MNBEDTW6ZDKYYBK6O7N.json","graph_json":"https://pith.science/api/pith-number/345KRE4MNBEDTW6ZDKYYBK6O7N/graph.json","events_json":"https://pith.science/api/pith-number/345KRE4MNBEDTW6ZDKYYBK6O7N/events.json","paper":"https://pith.science/paper/345KRE4M"},"agent_actions":{"view_html":"https://pith.science/pith/345KRE4MNBEDTW6ZDKYYBK6O7N","download_json":"https://pith.science/pith/345KRE4MNBEDTW6ZDKYYBK6O7N.json","view_paper":"https://pith.science/paper/345KRE4M","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.11628&json=true","fetch_graph":"https://pith.science/api/pith-number/345KRE4MNBEDTW6ZDKYYBK6O7N/graph.json","fetch_events":"https://pith.science/api/pith-number/345KRE4MNBEDTW6ZDKYYBK6O7N/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/345KRE4MNBEDTW6ZDKYYBK6O7N/action/timestamp_anchor","attest_storage":"https://pith.science/pith/345KRE4MNBEDTW6ZDKYYBK6O7N/action/storage_attestation","attest_author":"https://pith.science/pith/345KRE4MNBEDTW6ZDKYYBK6O7N/action/author_attestation","sign_citation":"https://pith.science/pith/345KRE4MNBEDTW6ZDKYYBK6O7N/action/citation_signature","submit_replication":"https://pith.science/pith/345KRE4MNBEDTW6ZDKYYBK6O7N/action/replication_record"}},"created_at":"2026-07-05T07:02:10.254641+00:00","updated_at":"2026-07-05T07:02:10.254641+00:00"}