{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:H7CWFKRWBD6T2IX6BF7MFJXHAV","short_pith_number":"pith:H7CWFKRW","schema_version":"1.0","canonical_sha256":"3fc562aa3608fd3d22fe097ec2a6e7055539c9031303812b55742d8c2d110a70","source":{"kind":"arxiv","id":"2412.03719","version":2},"attestation_state":"computed","paper":{"title":"From Language Models over Tokens to Language Models over Characters","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ben LeBrun, Brian DuSell, John Terilla, Juan Luis Gastaldi, Mario Giulianelli, Ryan Cotterell, Timothy J. O'Donnell, Tim Vieira","submitted_at":"2024-12-04T21:19:20Z","abstract_excerpt":"Modern language models are internally -- and mathematically -- distributions over $\\it{token}$ strings rather than $\\it{character}$ strings, posing numerous challenges for programmers building user applications on top of them. For example, if a prompt is specified as a character string, it must be tokenized before passing it to the token-level language model. Thus, the tokenizer and consequent processing are very sensitive to the specification of the prompt (e.g., whether the prompt ends with a space or not). This paper presents algorithms for converting token-level language models to characte"},"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":"2412.03719","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-04T21:19:20Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"b7cc6bd1590fc7f601dd4e6d1f43ce1acc6ee602fddf274ab2b4312820d87fd9","abstract_canon_sha256":"8ddd65d75f382e55e4c592a3430491368b0313032bfa13f02f7e6e5ff1345adc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:18:40.734485Z","signature_b64":"FxLE5O1b6ayKVwsZTfqYY+P5Hj9RTeIlhdjVQmyE5a5pYKUNTLbWGgygpFLqwjPloiOpxLa+uKY0gLWw5FdaBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3fc562aa3608fd3d22fe097ec2a6e7055539c9031303812b55742d8c2d110a70","last_reissued_at":"2026-07-05T11:18:40.733947Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:18:40.733947Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"From Language Models over Tokens to Language Models over Characters","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ben LeBrun, Brian DuSell, John Terilla, Juan Luis Gastaldi, Mario Giulianelli, Ryan Cotterell, Timothy J. O'Donnell, Tim Vieira","submitted_at":"2024-12-04T21:19:20Z","abstract_excerpt":"Modern language models are internally -- and mathematically -- distributions over $\\it{token}$ strings rather than $\\it{character}$ strings, posing numerous challenges for programmers building user applications on top of them. For example, if a prompt is specified as a character string, it must be tokenized before passing it to the token-level language model. Thus, the tokenizer and consequent processing are very sensitive to the specification of the prompt (e.g., whether the prompt ends with a space or not). This paper presents algorithms for converting token-level language models to characte"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.03719","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/2412.03719/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":"2412.03719","created_at":"2026-07-05T11:18:40.734002+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.03719v2","created_at":"2026-07-05T11:18:40.734002+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.03719","created_at":"2026-07-05T11:18:40.734002+00:00"},{"alias_kind":"pith_short_12","alias_value":"H7CWFKRWBD6T","created_at":"2026-07-05T11:18:40.734002+00:00"},{"alias_kind":"pith_short_16","alias_value":"H7CWFKRWBD6T2IX6","created_at":"2026-07-05T11:18:40.734002+00:00"},{"alias_kind":"pith_short_8","alias_value":"H7CWFKRW","created_at":"2026-07-05T11:18:40.734002+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2506.14123","citing_title":"Sampling from Your Language Model One Byte at a Time","ref_index":75,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/H7CWFKRWBD6T2IX6BF7MFJXHAV","json":"https://pith.science/pith/H7CWFKRWBD6T2IX6BF7MFJXHAV.json","graph_json":"https://pith.science/api/pith-number/H7CWFKRWBD6T2IX6BF7MFJXHAV/graph.json","events_json":"https://pith.science/api/pith-number/H7CWFKRWBD6T2IX6BF7MFJXHAV/events.json","paper":"https://pith.science/paper/H7CWFKRW"},"agent_actions":{"view_html":"https://pith.science/pith/H7CWFKRWBD6T2IX6BF7MFJXHAV","download_json":"https://pith.science/pith/H7CWFKRWBD6T2IX6BF7MFJXHAV.json","view_paper":"https://pith.science/paper/H7CWFKRW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.03719&json=true","fetch_graph":"https://pith.science/api/pith-number/H7CWFKRWBD6T2IX6BF7MFJXHAV/graph.json","fetch_events":"https://pith.science/api/pith-number/H7CWFKRWBD6T2IX6BF7MFJXHAV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/H7CWFKRWBD6T2IX6BF7MFJXHAV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/H7CWFKRWBD6T2IX6BF7MFJXHAV/action/storage_attestation","attest_author":"https://pith.science/pith/H7CWFKRWBD6T2IX6BF7MFJXHAV/action/author_attestation","sign_citation":"https://pith.science/pith/H7CWFKRWBD6T2IX6BF7MFJXHAV/action/citation_signature","submit_replication":"https://pith.science/pith/H7CWFKRWBD6T2IX6BF7MFJXHAV/action/replication_record"}},"created_at":"2026-07-05T11:18:40.734002+00:00","updated_at":"2026-07-05T11:18:40.734002+00:00"}