{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:HXMHXSWPBPE7PFX4VSNHRW3ISE","short_pith_number":"pith:HXMHXSWP","schema_version":"1.0","canonical_sha256":"3dd87bcacf0bc9f796fcac9a78db689133345832ee33b7a281ba06e1ac8623ea","source":{"kind":"arxiv","id":"1908.10322","version":1},"attestation_state":"computed","paper":{"title":"Bridging the Gap for Tokenizer-Free Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.IR","cs.LG"],"primary_cat":"cs.CL","authors_text":"Dokook Choe, Heeyoung Lee, Mandy Guo, Noah Constant, Rami Al-Rfou","submitted_at":"2019-08-27T16:53:59Z","abstract_excerpt":"Purely character-based language models (LMs) have been lagging in quality on large scale datasets, and current state-of-the-art LMs rely on word tokenization. It has been assumed that injecting the prior knowledge of a tokenizer into the model is essential to achieving competitive results. In this paper, we show that contrary to this conventional wisdom, tokenizer-free LMs with sufficient capacity can achieve competitive performance on a large scale dataset. We train a vanilla transformer network with 40 self-attention layers on the One Billion Word (lm1b) benchmark and achieve a new state of "},"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":"1908.10322","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-08-27T16:53:59Z","cross_cats_sorted":["cs.AI","cs.IR","cs.LG"],"title_canon_sha256":"5b89242964b904fbfdb0a63a34e5f97776ec455d5f6a461afb866177522e4b56","abstract_canon_sha256":"72bb734c4932d3d5cb4d289761cb7451016d2a4805329d9dfa90c79e7f38d64e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:00:10.973279Z","signature_b64":"hNbsl9WDkFLXIUnbbwbf2x6YKK0lXZaKwQ3Wncy+9erEihxVKUM3H5ef0vcAViFns5wuTsl1yRqjhJbQn5scCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3dd87bcacf0bc9f796fcac9a78db689133345832ee33b7a281ba06e1ac8623ea","last_reissued_at":"2026-07-05T00:00:10.972877Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:00:10.972877Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Bridging the Gap for Tokenizer-Free Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.IR","cs.LG"],"primary_cat":"cs.CL","authors_text":"Dokook Choe, Heeyoung Lee, Mandy Guo, Noah Constant, Rami Al-Rfou","submitted_at":"2019-08-27T16:53:59Z","abstract_excerpt":"Purely character-based language models (LMs) have been lagging in quality on large scale datasets, and current state-of-the-art LMs rely on word tokenization. It has been assumed that injecting the prior knowledge of a tokenizer into the model is essential to achieving competitive results. In this paper, we show that contrary to this conventional wisdom, tokenizer-free LMs with sufficient capacity can achieve competitive performance on a large scale dataset. We train a vanilla transformer network with 40 self-attention layers on the One Billion Word (lm1b) benchmark and achieve a new state of "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1908.10322","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/1908.10322/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":"1908.10322","created_at":"2026-07-05T00:00:10.972934+00:00"},{"alias_kind":"arxiv_version","alias_value":"1908.10322v1","created_at":"2026-07-05T00:00:10.972934+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1908.10322","created_at":"2026-07-05T00:00:10.972934+00:00"},{"alias_kind":"pith_short_12","alias_value":"HXMHXSWPBPE7","created_at":"2026-07-05T00:00:10.972934+00:00"},{"alias_kind":"pith_short_16","alias_value":"HXMHXSWPBPE7PFX4","created_at":"2026-07-05T00:00:10.972934+00:00"},{"alias_kind":"pith_short_8","alias_value":"HXMHXSWP","created_at":"2026-07-05T00:00:10.972934+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.09630","citing_title":"Scratchpad Patching: Decoupling Compute from Patch Size in Byte-Level Language Models","ref_index":15,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HXMHXSWPBPE7PFX4VSNHRW3ISE","json":"https://pith.science/pith/HXMHXSWPBPE7PFX4VSNHRW3ISE.json","graph_json":"https://pith.science/api/pith-number/HXMHXSWPBPE7PFX4VSNHRW3ISE/graph.json","events_json":"https://pith.science/api/pith-number/HXMHXSWPBPE7PFX4VSNHRW3ISE/events.json","paper":"https://pith.science/paper/HXMHXSWP"},"agent_actions":{"view_html":"https://pith.science/pith/HXMHXSWPBPE7PFX4VSNHRW3ISE","download_json":"https://pith.science/pith/HXMHXSWPBPE7PFX4VSNHRW3ISE.json","view_paper":"https://pith.science/paper/HXMHXSWP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1908.10322&json=true","fetch_graph":"https://pith.science/api/pith-number/HXMHXSWPBPE7PFX4VSNHRW3ISE/graph.json","fetch_events":"https://pith.science/api/pith-number/HXMHXSWPBPE7PFX4VSNHRW3ISE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HXMHXSWPBPE7PFX4VSNHRW3ISE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HXMHXSWPBPE7PFX4VSNHRW3ISE/action/storage_attestation","attest_author":"https://pith.science/pith/HXMHXSWPBPE7PFX4VSNHRW3ISE/action/author_attestation","sign_citation":"https://pith.science/pith/HXMHXSWPBPE7PFX4VSNHRW3ISE/action/citation_signature","submit_replication":"https://pith.science/pith/HXMHXSWPBPE7PFX4VSNHRW3ISE/action/replication_record"}},"created_at":"2026-07-05T00:00:10.972934+00:00","updated_at":"2026-07-05T00:00:10.972934+00:00"}