{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:JRANJ5GWYB7XNPMQLE3SLXQPPQ","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":"ccaf65be5eef0205d8c1f05f1ff5c1cd324209ad675a555b4470308eb6c3b59e","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-09-06T20:12:34Z","title_canon_sha256":"1aaf8f9475b9f94caaba630567e6bdab4b07d57d807bef0945d73d64a34db1f7"},"schema_version":"1.0","source":{"id":"2409.04599","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2409.04599","created_at":"2026-07-05T09:04:15Z"},{"alias_kind":"arxiv_version","alias_value":"2409.04599v1","created_at":"2026-07-05T09:04:15Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.04599","created_at":"2026-07-05T09:04:15Z"},{"alias_kind":"pith_short_12","alias_value":"JRANJ5GWYB7X","created_at":"2026-07-05T09:04:15Z"},{"alias_kind":"pith_short_16","alias_value":"JRANJ5GWYB7XNPMQ","created_at":"2026-07-05T09:04:15Z"},{"alias_kind":"pith_short_8","alias_value":"JRANJ5GW","created_at":"2026-07-05T09:04:15Z"}],"graph_snapshots":[{"event_id":"sha256:abe7f803464ef120baccdd57fcd56381781fa3553de7a91243246e1153dfe145","target":"graph","created_at":"2026-07-05T09:04:15Z","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/2409.04599/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Language models can largely benefit from efficient tokenization. However, they still mostly utilize the classical BPE algorithm, a simple and reliable method. This has been shown to cause such issues as under-trained tokens and sub-optimal compression that may affect the downstream performance. We introduce Picky BPE, a modified BPE algorithm that carries out vocabulary refinement during tokenizer training. Our method improves vocabulary efficiency, eliminates under-trained tokens, and does not compromise text compression. Our experiments show that our method does not reduce the downstream per","authors_text":"Catherine Arnett, Elizaveta Korotkova, Ivan P. Yamshchikov, Pavel Chizhov","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-09-06T20:12:34Z","title":"BPE Gets Picky: Efficient Vocabulary Refinement During Tokenizer Training"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.04599","kind":"arxiv","version":1},"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":"sha256:3737ad77db2e51579b4315454030a2825526ffe86f9bad199de261946da3598f","target":"record","created_at":"2026-07-05T09:04:15Z","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":"ccaf65be5eef0205d8c1f05f1ff5c1cd324209ad675a555b4470308eb6c3b59e","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-09-06T20:12:34Z","title_canon_sha256":"1aaf8f9475b9f94caaba630567e6bdab4b07d57d807bef0945d73d64a34db1f7"},"schema_version":"1.0","source":{"id":"2409.04599","kind":"arxiv","version":1}},"canonical_sha256":"4c40d4f4d6c07f76bd90593725de0f7c0e88df8a47fa198615e77c467b0146de","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4c40d4f4d6c07f76bd90593725de0f7c0e88df8a47fa198615e77c467b0146de","first_computed_at":"2026-07-05T09:04:15.010466Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:04:15.010466Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"367+/n3Ovjtgo12jmLl1YS3StHavoQeEWFXTCmRv9yZ3I/TJu5R2jO66gOKxk8FWlz9d8P1PA3EhanSrvEGCCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T09:04:15.010873Z","signed_message":"canonical_sha256_bytes"},"source_id":"2409.04599","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:3737ad77db2e51579b4315454030a2825526ffe86f9bad199de261946da3598f","sha256:abe7f803464ef120baccdd57fcd56381781fa3553de7a91243246e1153dfe145"],"state_sha256":"dc3ce1232d07551039167765c994130dfcc73e3bb90930d5d3590224f5d3cf8e"}