Glitch Tokens in Large Language Models: Categorization Taxonomy and Effective Detection
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866913321972662272 |
|---|---|
| author | Li, Yuxi Liu, Yi Deng, Gelei Zhang, Ying Song, Wenjia Shi, Ling Wang, Kailong Li, Yuekang Liu, Yang Wang, Haoyu |
| author_facet | Li, Yuxi Liu, Yi Deng, Gelei Zhang, Ying Song, Wenjia Shi, Ling Wang, Kailong Li, Yuekang Liu, Yang Wang, Haoyu |
| contents | With the expanding application of Large Language Models (LLMs) in various domains, it becomes imperative to comprehensively investigate their unforeseen behaviors and consequent outcomes. In this study, we introduce and systematically explore the phenomenon of "glitch tokens", which are anomalous tokens produced by established tokenizers and could potentially compromise the models' quality of response. Specifically, we experiment on seven top popular LLMs utilizing three distinct tokenizers and involving a totally of 182,517 tokens. We present categorizations of the identified glitch tokens and symptoms exhibited by LLMs when interacting with glitch tokens. Based on our observation that glitch tokens tend to cluster in the embedding space, we propose GlitchHunter, a novel iterative clustering-based technique, for efficient glitch token detection. The evaluation shows that our approach notably outperforms three baseline methods on eight open-source LLMs. To the best of our knowledge, we present the first comprehensive study on glitch tokens. Our new detection further provides valuable insights into mitigating tokenization-related errors in LLMs. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_09894 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Glitch Tokens in Large Language Models: Categorization Taxonomy and Effective Detection Li, Yuxi Liu, Yi Deng, Gelei Zhang, Ying Song, Wenjia Shi, Ling Wang, Kailong Li, Yuekang Liu, Yang Wang, Haoyu Computation and Language Software Engineering With the expanding application of Large Language Models (LLMs) in various domains, it becomes imperative to comprehensively investigate their unforeseen behaviors and consequent outcomes. In this study, we introduce and systematically explore the phenomenon of "glitch tokens", which are anomalous tokens produced by established tokenizers and could potentially compromise the models' quality of response. Specifically, we experiment on seven top popular LLMs utilizing three distinct tokenizers and involving a totally of 182,517 tokens. We present categorizations of the identified glitch tokens and symptoms exhibited by LLMs when interacting with glitch tokens. Based on our observation that glitch tokens tend to cluster in the embedding space, we propose GlitchHunter, a novel iterative clustering-based technique, for efficient glitch token detection. The evaluation shows that our approach notably outperforms three baseline methods on eight open-source LLMs. To the best of our knowledge, we present the first comprehensive study on glitch tokens. Our new detection further provides valuable insights into mitigating tokenization-related errors in LLMs. |
| title | Glitch Tokens in Large Language Models: Categorization Taxonomy and Effective Detection |
| topic | Computation and Language Software Engineering |
| url | https://arxiv.org/abs/2404.09894 |