{"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/t-free-tokenizer-free-generative-llms-via","title":"T-FREE: Subword Tokenizer-Free Generative LLMs via Sparse Representations for Memory-Efficient Embeddings","arxiv_id":"2406.19223","date":"2024-06-27","proceeding":null,"authors":["Björn Deiseroth","Manuel Brack","Patrick Schramowski","Kristian Kersting","Samuel Weinbach"],"abstract":"Tokenizers are crucial for encoding information in Large Language Models, but their development has recently stagnated, and they contain inherent weaknesses. Major limitations include computational overhead, ineffective vocabulary use, and unnecessarily large embedding and head layers. Additionally, their performance is biased towards a reference corpus, leading to reduced effectiveness for underrepresented languages. To remedy these issues, we propose T-FREE, which directly embeds words through sparse activation patterns over character triplets, and does not require a reference corpus. T-FREE inherently exploits morphological similarities and allows for strong compression of embedding layers. In our exhaustive experimental evaluation, we achieve competitive downstream performance with a parameter reduction of more than 85% on these layers. Further, T-FREE shows significant improvements in cross-lingual transfer learning.","url_abs":"https://arxiv.org/abs/2406.19223v2","url_pdf":"https://arxiv.org/pdf/2406.19223v2.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":"t-free-tokenizer-free-generative-llms-via","repo_url":"https://github.com/aleph-alpha/trigrams","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"NOASSERTION"}}],"tasks":[{"task_slug":"cross-lingual-transfer","task_name":"Cross-Lingual Transfer"},{"task_slug":"transfer-learning","task_name":"Transfer Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2406.19223","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2406.19223"}},"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/aleph-alpha/trigrams","reach":{"status":"ok","spdx":"NOASSERTION"}}],"summary":{"ran":2,"unverified":2},"by_repo_kind":{"official":{"samples":4,"ran":2,"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":4,"samples":[{"code_sha256_prefix":"951454bdf77e2b32","entry":"command_exists","repo":"aleph-alpha/trigrams","repo_kind":"official","path":"src/trigram_tokenizer/logging/logging.py","file_url":"https://github.com/aleph-alpha/trigrams/blob/HEAD/src/trigram_tokenizer/logging/logging.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"951454bdf77e2b32"}},{"code_sha256_prefix":"09086ca4d0ba2f20","entry":"is_date","repo":"aleph-alpha/trigrams","repo_kind":"official","path":"src/trigram_tokenizer/logging/logging.py","file_url":"https://github.com/aleph-alpha/trigrams/blob/HEAD/src/trigram_tokenizer/logging/logging.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"09086ca4d0ba2f20"}},{"code_sha256_prefix":"a09a9eadb25726cb","entry":"allreduce_bucket","repo":"aleph-alpha/trigrams","repo_kind":"official","path":"src/trigram_tokenizer/optimizer/allreduce.py","file_url":"https://github.com/aleph-alpha/trigrams/blob/HEAD/src/trigram_tokenizer/optimizer/allreduce.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"a09a9eadb25726cb"}},{"code_sha256_prefix":"afb6812914d4779d","entry":"split_dtypes","repo":"aleph-alpha/trigrams","repo_kind":"official","path":"src/trigram_tokenizer/optimizer/allreduce.py","file_url":"https://github.com/aleph-alpha/trigrams/blob/HEAD/src/trigram_tokenizer/optimizer/allreduce.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"afb6812914d4779d"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}