{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:J2AD5BU2PFULNZS72S2R3XBQAM","short_pith_number":"pith:J2AD5BU2","canonical_record":{"source":{"id":"2404.06508","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-09T17:57:29Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"d87c50cf79e2dbb7bc6bfa666934123974e24a5207e48958972620bfc21d069f","abstract_canon_sha256":"cf7572af735c4342e10fd3066bfd30a2c4b6fb140c45054caca1f7aa02ffe168"},"schema_version":"1.0"},"canonical_sha256":"4e803e869a7968b6e65fd4b51ddc30033b8114e39897970819c9236623a405fb","source":{"kind":"arxiv","id":"2404.06508","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2404.06508","created_at":"2026-07-05T08:44:53Z"},{"alias_kind":"arxiv_version","alias_value":"2404.06508v3","created_at":"2026-07-05T08:44:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.06508","created_at":"2026-07-05T08:44:53Z"},{"alias_kind":"pith_short_12","alias_value":"J2AD5BU2PFUL","created_at":"2026-07-05T08:44:53Z"},{"alias_kind":"pith_short_16","alias_value":"J2AD5BU2PFULNZS7","created_at":"2026-07-05T08:44:53Z"},{"alias_kind":"pith_short_8","alias_value":"J2AD5BU2","created_at":"2026-07-05T08:44:53Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:J2AD5BU2PFULNZS72S2R3XBQAM","target":"record","payload":{"canonical_record":{"source":{"id":"2404.06508","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-09T17:57:29Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"d87c50cf79e2dbb7bc6bfa666934123974e24a5207e48958972620bfc21d069f","abstract_canon_sha256":"cf7572af735c4342e10fd3066bfd30a2c4b6fb140c45054caca1f7aa02ffe168"},"schema_version":"1.0"},"canonical_sha256":"4e803e869a7968b6e65fd4b51ddc30033b8114e39897970819c9236623a405fb","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:44:53.344636Z","signature_b64":"JAP+jlUoVZF+JVZTu7dwQTZG4GVybzScInQ2+LkstxQ2lTHHikEvVuxZpyE0lTfLXE6Pxdej61BsWsKXhphnDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4e803e869a7968b6e65fd4b51ddc30033b8114e39897970819c9236623a405fb","last_reissued_at":"2026-07-05T08:44:53.344076Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:44:53.344076Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2404.06508","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":"2026-07-05T08:44:53Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"bfDw+jFOlh/Nmx2YJGrE1iBxGaDCTD30julxYqQg22pi3pAJeWShZniN+eiT+PWy5ns5H8RwlaruipI8DFH+Cg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T11:28:22.381060Z"},"content_sha256":"c7b6d3aaac39dba4d22b9331465fc2db1539b6fb8b29681effea1516ff0a4a47","schema_version":"1.0","event_id":"sha256:c7b6d3aaac39dba4d22b9331465fc2db1539b6fb8b29681effea1516ff0a4a47"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:J2AD5BU2PFULNZS72S2R3XBQAM","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"On the Effect of (Near) Duplicate Subwords in Language Modelling","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Anton Sch\\\"afer, Imanol Schlag, Thomas Hofmann, Tiago Pimentel","submitted_at":"2024-04-09T17:57:29Z","abstract_excerpt":"Tokenisation is a core part of language models (LMs). It involves splitting a character sequence into subwords which are assigned arbitrary indices before being served to the LM. While typically lossless, however, this process may lead to less sample efficient LM training: as it removes character-level information, it could make it harder for LMs to generalise across similar subwords, such as now and Now. We refer to such subwords as near duplicates. In this paper, we study the impact of near duplicate subwords on LM training efficiency. First, we design an experiment that gives us an upper bo"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.06508","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/2404.06508/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":"2026-07-05T08:44:53Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"urmXT9aU3Qmk/epq2uisgHASQLCGHgs+x2GZ4IBnIltC1S/7TEiQGNNIkGuyK14CeKsdc8hw49/pejY8AL0MDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T11:28:22.381650Z"},"content_sha256":"307abb72f776406ce612a6a09a79b96ebe2f750e0d12ad1549b769a08d023ef7","schema_version":"1.0","event_id":"sha256:307abb72f776406ce612a6a09a79b96ebe2f750e0d12ad1549b769a08d023ef7"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/J2AD5BU2PFULNZS72S2R3XBQAM/bundle.json","state_url":"https://pith.science/pith/J2AD5BU2PFULNZS72S2R3XBQAM/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/J2AD5BU2PFULNZS72S2R3XBQAM/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":"2026-08-06T11:28:22Z","links":{"resolver":"https://pith.science/pith/J2AD5BU2PFULNZS72S2R3XBQAM","bundle":"https://pith.science/pith/J2AD5BU2PFULNZS72S2R3XBQAM/bundle.json","state":"https://pith.science/pith/J2AD5BU2PFULNZS72S2R3XBQAM/state.json","well_known_bundle":"https://pith.science/.well-known/pith/J2AD5BU2PFULNZS72S2R3XBQAM/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:J2AD5BU2PFULNZS72S2R3XBQAM","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":"cf7572af735c4342e10fd3066bfd30a2c4b6fb140c45054caca1f7aa02ffe168","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-09T17:57:29Z","title_canon_sha256":"d87c50cf79e2dbb7bc6bfa666934123974e24a5207e48958972620bfc21d069f"},"schema_version":"1.0","source":{"id":"2404.06508","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2404.06508","created_at":"2026-07-05T08:44:53Z"},{"alias_kind":"arxiv_version","alias_value":"2404.06508v3","created_at":"2026-07-05T08:44:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.06508","created_at":"2026-07-05T08:44:53Z"},{"alias_kind":"pith_short_12","alias_value":"J2AD5BU2PFUL","created_at":"2026-07-05T08:44:53Z"},{"alias_kind":"pith_short_16","alias_value":"J2AD5BU2PFULNZS7","created_at":"2026-07-05T08:44:53Z"},{"alias_kind":"pith_short_8","alias_value":"J2AD5BU2","created_at":"2026-07-05T08:44:53Z"}],"graph_snapshots":[{"event_id":"sha256:307abb72f776406ce612a6a09a79b96ebe2f750e0d12ad1549b769a08d023ef7","target":"graph","created_at":"2026-07-05T08:44:53Z","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/2404.06508/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Tokenisation is a core part of language models (LMs). It involves splitting a character sequence into subwords which are assigned arbitrary indices before being served to the LM. While typically lossless, however, this process may lead to less sample efficient LM training: as it removes character-level information, it could make it harder for LMs to generalise across similar subwords, such as now and Now. We refer to such subwords as near duplicates. In this paper, we study the impact of near duplicate subwords on LM training efficiency. First, we design an experiment that gives us an upper bo","authors_text":"Anton Sch\\\"afer, Imanol Schlag, Thomas Hofmann, Tiago Pimentel","cross_cats":["cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-09T17:57:29Z","title":"On the Effect of (Near) Duplicate Subwords in Language Modelling"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.06508","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:c7b6d3aaac39dba4d22b9331465fc2db1539b6fb8b29681effea1516ff0a4a47","target":"record","created_at":"2026-07-05T08:44:53Z","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":"cf7572af735c4342e10fd3066bfd30a2c4b6fb140c45054caca1f7aa02ffe168","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-09T17:57:29Z","title_canon_sha256":"d87c50cf79e2dbb7bc6bfa666934123974e24a5207e48958972620bfc21d069f"},"schema_version":"1.0","source":{"id":"2404.06508","kind":"arxiv","version":3}},"canonical_sha256":"4e803e869a7968b6e65fd4b51ddc30033b8114e39897970819c9236623a405fb","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4e803e869a7968b6e65fd4b51ddc30033b8114e39897970819c9236623a405fb","first_computed_at":"2026-07-05T08:44:53.344076Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:44:53.344076Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"JAP+jlUoVZF+JVZTu7dwQTZG4GVybzScInQ2+LkstxQ2lTHHikEvVuxZpyE0lTfLXE6Pxdej61BsWsKXhphnDQ==","signature_status":"signed_v1","signed_at":"2026-07-05T08:44:53.344636Z","signed_message":"canonical_sha256_bytes"},"source_id":"2404.06508","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c7b6d3aaac39dba4d22b9331465fc2db1539b6fb8b29681effea1516ff0a4a47","sha256:307abb72f776406ce612a6a09a79b96ebe2f750e0d12ad1549b769a08d023ef7"],"state_sha256":"2d5c4cd7e9fdaff27c428bd905bc34cc203355a96f70f6bb171007f905484c3d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"JyJyjy6/8ze50w9/KNjsi2M49rPsYXt+fjyWwbZxwcReThjhAEm28T+JkijHAfUnRQf/UyE2MARpKKQ6hl5TBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-06T11:28:22.387772Z","bundle_sha256":"f0c5490d68a4f4f9be9104f4738e1d070924ac588a6471f9121f618059d016fb"}}