{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2020:YQXA5CKTSKXW5IK7DJEHEPOXQA","short_pith_number":"pith:YQXA5CKT","canonical_record":{"source":{"id":"2012.15524","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-12-31T10:01:29Z","cross_cats_sorted":[],"title_canon_sha256":"f63178ae4b76b1af4ca92e76a4453e6f52bab295c5b8528faf7e3d02e7b336a1","abstract_canon_sha256":"2cb6fc0aa705af632cdfb46b943c8dde9bbe5da99344c9aaaaefc8c31ebb6c52"},"schema_version":"1.0"},"canonical_sha256":"c42e0e895392af6ea15f1a48723dd78016fb7c234e45c945dc191e274c6fbdd7","source":{"kind":"arxiv","id":"2012.15524","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2012.15524","created_at":"2026-07-05T03:20:19Z"},{"alias_kind":"arxiv_version","alias_value":"2012.15524v3","created_at":"2026-07-05T03:20:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2012.15524","created_at":"2026-07-05T03:20:19Z"},{"alias_kind":"pith_short_12","alias_value":"YQXA5CKTSKXW","created_at":"2026-07-05T03:20:19Z"},{"alias_kind":"pith_short_16","alias_value":"YQXA5CKTSKXW5IK7","created_at":"2026-07-05T03:20:19Z"},{"alias_kind":"pith_short_8","alias_value":"YQXA5CKT","created_at":"2026-07-05T03:20:19Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2020:YQXA5CKTSKXW5IK7DJEHEPOXQA","target":"record","payload":{"canonical_record":{"source":{"id":"2012.15524","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-12-31T10:01:29Z","cross_cats_sorted":[],"title_canon_sha256":"f63178ae4b76b1af4ca92e76a4453e6f52bab295c5b8528faf7e3d02e7b336a1","abstract_canon_sha256":"2cb6fc0aa705af632cdfb46b943c8dde9bbe5da99344c9aaaaefc8c31ebb6c52"},"schema_version":"1.0"},"canonical_sha256":"c42e0e895392af6ea15f1a48723dd78016fb7c234e45c945dc191e274c6fbdd7","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:20:19.448583Z","signature_b64":"CxsTX2VHE/Oi7PT4/6Yvad4nw1GoyWFhxaW0zhMNL1c2fiPNASK6j7VQQDlxXHNmQWzOwlQDb/4yS50fa5h8Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c42e0e895392af6ea15f1a48723dd78016fb7c234e45c945dc191e274c6fbdd7","last_reissued_at":"2026-07-05T03:20:19.448191Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:20:19.448191Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2012.15524","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T03:20:19Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"wsMY1hDAcmpYsx9fsBaCwCRJME3UIhB1bppQrEjMEgEaa87BYiJR9cFrRnkYgkaMVrT5v8Pmc3CoqXkUQciWDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T11:32:15.066086Z"},"content_sha256":"13eb2dc28a8305c7f1ef4593af7aae5116b77a9afb4f3d5f7d5a3af16bab1b7f","schema_version":"1.0","event_id":"sha256:13eb2dc28a8305c7f1ef4593af7aae5116b77a9afb4f3d5f7d5a3af16bab1b7f"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2020:YQXA5CKTSKXW5IK7DJEHEPOXQA","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Fast WordPiece Tokenization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alex Salcianu, Dave Dopson, Denny Zhou, Xinying Song, Yang Song","submitted_at":"2020-12-31T10:01:29Z","abstract_excerpt":"Tokenization is a fundamental preprocessing step for almost all NLP tasks. In this paper, we propose efficient algorithms for the WordPiece tokenization used in BERT, from single-word tokenization to general text (e.g., sentence) tokenization. When tokenizing a single word, WordPiece uses a longest-match-first strategy, known as maximum matching. The best known algorithms so far are O(n^2) (where n is the input length) or O(nm) (where m is the maximum vocabulary token length). We propose a novel algorithm whose tokenization complexity is strictly O(n). Our method is inspired by the Aho-Corasic"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2012.15524","kind":"arxiv","version":3},"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/2012.15524/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T03:20:19Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"gakVxlyb6rwjPvuzJTU6F/1ob9CO4RB5MufcsHwlktATVqXrTqZqVj3ltlXYXDNMJDaqIRmb+t3CIT3g4b1NBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T11:32:15.067316Z"},"content_sha256":"7f9db39c361faf916e6ba342e406e8d7332d3246a366453b5554dd3d4268db46","schema_version":"1.0","event_id":"sha256:7f9db39c361faf916e6ba342e406e8d7332d3246a366453b5554dd3d4268db46"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/YQXA5CKTSKXW5IK7DJEHEPOXQA/bundle.json","state_url":"https://pith.science/pith/YQXA5CKTSKXW5IK7DJEHEPOXQA/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/YQXA5CKTSKXW5IK7DJEHEPOXQA/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-10T11:32:15Z","links":{"resolver":"https://pith.science/pith/YQXA5CKTSKXW5IK7DJEHEPOXQA","bundle":"https://pith.science/pith/YQXA5CKTSKXW5IK7DJEHEPOXQA/bundle.json","state":"https://pith.science/pith/YQXA5CKTSKXW5IK7DJEHEPOXQA/state.json","well_known_bundle":"https://pith.science/.well-known/pith/YQXA5CKTSKXW5IK7DJEHEPOXQA/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2020:YQXA5CKTSKXW5IK7DJEHEPOXQA","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"2cb6fc0aa705af632cdfb46b943c8dde9bbe5da99344c9aaaaefc8c31ebb6c52","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-12-31T10:01:29Z","title_canon_sha256":"f63178ae4b76b1af4ca92e76a4453e6f52bab295c5b8528faf7e3d02e7b336a1"},"schema_version":"1.0","source":{"id":"2012.15524","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2012.15524","created_at":"2026-07-05T03:20:19Z"},{"alias_kind":"arxiv_version","alias_value":"2012.15524v3","created_at":"2026-07-05T03:20:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2012.15524","created_at":"2026-07-05T03:20:19Z"},{"alias_kind":"pith_short_12","alias_value":"YQXA5CKTSKXW","created_at":"2026-07-05T03:20:19Z"},{"alias_kind":"pith_short_16","alias_value":"YQXA5CKTSKXW5IK7","created_at":"2026-07-05T03:20:19Z"},{"alias_kind":"pith_short_8","alias_value":"YQXA5CKT","created_at":"2026-07-05T03:20:19Z"}],"graph_snapshots":[{"event_id":"sha256:7f9db39c361faf916e6ba342e406e8d7332d3246a366453b5554dd3d4268db46","target":"graph","created_at":"2026-07-05T03:20:19Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2012.15524/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Tokenization is a fundamental preprocessing step for almost all NLP tasks. In this paper, we propose efficient algorithms for the WordPiece tokenization used in BERT, from single-word tokenization to general text (e.g., sentence) tokenization. When tokenizing a single word, WordPiece uses a longest-match-first strategy, known as maximum matching. The best known algorithms so far are O(n^2) (where n is the input length) or O(nm) (where m is the maximum vocabulary token length). We propose a novel algorithm whose tokenization complexity is strictly O(n). Our method is inspired by the Aho-Corasic","authors_text":"Alex Salcianu, Dave Dopson, Denny Zhou, Xinying Song, Yang Song","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-12-31T10:01:29Z","title":"Fast WordPiece Tokenization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2012.15524","kind":"arxiv","version":3},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:13eb2dc28a8305c7f1ef4593af7aae5116b77a9afb4f3d5f7d5a3af16bab1b7f","target":"record","created_at":"2026-07-05T03:20:19Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"2cb6fc0aa705af632cdfb46b943c8dde9bbe5da99344c9aaaaefc8c31ebb6c52","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-12-31T10:01:29Z","title_canon_sha256":"f63178ae4b76b1af4ca92e76a4453e6f52bab295c5b8528faf7e3d02e7b336a1"},"schema_version":"1.0","source":{"id":"2012.15524","kind":"arxiv","version":3}},"canonical_sha256":"c42e0e895392af6ea15f1a48723dd78016fb7c234e45c945dc191e274c6fbdd7","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c42e0e895392af6ea15f1a48723dd78016fb7c234e45c945dc191e274c6fbdd7","first_computed_at":"2026-07-05T03:20:19.448191Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T03:20:19.448191Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"CxsTX2VHE/Oi7PT4/6Yvad4nw1GoyWFhxaW0zhMNL1c2fiPNASK6j7VQQDlxXHNmQWzOwlQDb/4yS50fa5h8Dg==","signature_status":"signed_v1","signed_at":"2026-07-05T03:20:19.448583Z","signed_message":"canonical_sha256_bytes"},"source_id":"2012.15524","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:13eb2dc28a8305c7f1ef4593af7aae5116b77a9afb4f3d5f7d5a3af16bab1b7f","sha256:7f9db39c361faf916e6ba342e406e8d7332d3246a366453b5554dd3d4268db46"],"state_sha256":"db3dd7ee44cfa7d6671e2ba222ef52f892086a9dc9992869ae90a8c1f4f1d1dc"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"gMcEXOMZbYNTpTXd6NBRoDnRKKXuP/pE5P5sMWQEwrUH1iAafRBR6CbQpbzk7zrzd6daBrDvzA+8jqTaWNuDBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-10T11:32:15.075503Z","bundle_sha256":"6ecc8529b892e200ccab52a6b492279d0d105188f84f55560ccd711c968cc47c"}}