{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/bpe-stays-on-script-structured-encoding-for","title":"BPE Stays on SCRIPT: Structured Encoding for Robust Multilingual Pretokenization","arxiv_id":"2505.24689","date":"2025-05-30","proceeding":null,"authors":["Sander Land","Catherine Arnett"],"abstract":"Byte Pair Encoding (BPE) tokenizers, widely used in Large Language Models, face challenges in multilingual settings, including penalization of non-Western scripts and the creation of tokens with partial UTF-8 sequences. Pretokenization, often reliant on complex regular expressions, can also introduce fragility and unexpected edge cases. We propose SCRIPT (Script Category Representation in PreTokenization), a novel encoding scheme that bypasses UTF-8 byte conversion by using initial tokens based on Unicode script and category properties. This approach enables a simple, rule-based pretokenization strategy that respects script boundaries, offering a robust alternative to pretokenization strategies based on regular expressions. We also introduce and validate a constrained BPE merging strategy that enforces character integrity, applicable to both SCRIPT-BPE and byte-based BPE. Our experiments demonstrate that SCRIPT-BPE achieves competitive compression while eliminating encoding-based penalties for non-Latin-script languages.","url_abs":"https://arxiv.org/abs/2505.24689v1","url_pdf":"https://arxiv.org/pdf/2505.24689v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"bpe-stays-on-script-structured-encoding-for","repo_url":"https://github.com/sanderland/script_bpe","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[],"methods":[{"method_slug":"bpe","method_name":"BPE"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2505.24689","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2505.24689"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/sanderland/script_bpe","reach":{"status":"ok","spdx":"Apache-2.0"}}],"summary":{"ran":12,"unverified":1},"by_repo_kind":{"official":{"samples":13,"ran":12,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"e87d24f9415e5231","entry":"char_name","repo":"sanderland/script_bpe","repo_kind":"official","path":"script_bpe/pretokenize/scriptencoding.py","file_url":"https://github.com/sanderland/script_bpe/blob/HEAD/script_bpe/pretokenize/scriptencoding.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"e87d24f9415e5231"}},{"code_sha256_prefix":"596259fa949215c0","entry":"flatten_model_metadata","repo":"sanderland/script_bpe","repo_kind":"official","path":"script_bpe/analysis/experiments.py","file_url":"https://github.com/sanderland/script_bpe/blob/HEAD/script_bpe/analysis/experiments.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"596259fa949215c0"}},{"code_sha256_prefix":"bd0a8fb03ae8a736","entry":"format_latex_value","repo":"sanderland/script_bpe","repo_kind":"official","path":"script_bpe/analysis/formatting.py","file_url":"https://github.com/sanderland/script_bpe/blob/HEAD/script_bpe/analysis/formatting.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"bd0a8fb03ae8a736"}},{"code_sha256_prefix":"7a5f392192c6a1f1","entry":"format_results","repo":"sanderland/script_bpe","repo_kind":"official","path":"paper_utils/script_bpe/monolingual_compression.py","file_url":"https://github.com/sanderland/script_bpe/blob/HEAD/paper_utils/script_bpe/monolingual_compression.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"7a5f392192c6a1f1"}},{"code_sha256_prefix":"bb57f9ebf56c9d71","entry":"format_tokens_millions","repo":"sanderland/script_bpe","repo_kind":"official","path":"script_bpe/analysis/formatting.py","file_url":"https://github.com/sanderland/script_bpe/blob/HEAD/script_bpe/analysis/formatting.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"bb57f9ebf56c9d71"}},{"code_sha256_prefix":"e73bf9272554cac3","entry":"format_with_relchange","repo":"sanderland/script_bpe","repo_kind":"official","path":"script_bpe/analysis/formatting.py","file_url":"https://github.com/sanderland/script_bpe/blob/HEAD/script_bpe/analysis/formatting.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"e73bf9272554cac3"}},{"code_sha256_prefix":"75ffe04bfd9ceece","entry":"get_config_hash","repo":"sanderland/script_bpe","repo_kind":"official","path":"script_bpe/analysis/experiments.py","file_url":"https://github.com/sanderland/script_bpe/blob/HEAD/script_bpe/analysis/experiments.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"75ffe04bfd9ceece"}},{"code_sha256_prefix":"3654bb76cf003e14","entry":"join_workers","repo":"sanderland/script_bpe","repo_kind":"official","path":"script_bpe/utils.py","file_url":"https://github.com/sanderland/script_bpe/blob/HEAD/script_bpe/utils.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"3654bb76cf003e14"}},{"code_sha256_prefix":"0d64647d6f44ca2c","entry":"normalize_whitespace","repo":"sanderland/script_bpe","repo_kind":"official","path":"script_bpe/corpus/registry.py","file_url":"https://github.com/sanderland/script_bpe/blob/HEAD/script_bpe/corpus/registry.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"0d64647d6f44ca2c"}},{"code_sha256_prefix":"c60aff84445cbfba","entry":"select_drop_batch","repo":"sanderland/script_bpe","repo_kind":"official","path":"script_bpe/tokenizers/_mi_prune.py","file_url":"https://github.com/sanderland/script_bpe/blob/HEAD/script_bpe/tokenizers/_mi_prune.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"c60aff84445cbfba"}},{"code_sha256_prefix":"59a3a3ed2eef89cb","entry":"shutdown_pool","repo":"sanderland/script_bpe","repo_kind":"official","path":"script_bpe/utils.py","file_url":"https://github.com/sanderland/script_bpe/blob/HEAD/script_bpe/utils.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"59a3a3ed2eef89cb"}},{"code_sha256_prefix":"2b40f6e6aaa2fd22","entry":"token_array","repo":"sanderland/script_bpe","repo_kind":"official","path":"script_bpe/utils.py","file_url":"https://github.com/sanderland/script_bpe/blob/HEAD/script_bpe/utils.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"2b40f6e6aaa2fd22"}},{"code_sha256_prefix":"b946980d96fb85ed","entry":"compute_mi_table","repo":"sanderland/script_bpe","repo_kind":"official","path":"script_bpe/tokenizers/_mi_prune.py","file_url":"https://github.com/sanderland/script_bpe/blob/HEAD/script_bpe/tokenizers/_mi_prune.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"b946980d96fb85ed"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}