Revisiting subword tokenization: A case study on affixal negation in large language models
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866913298395430912 |
|---|---|
| author | Truong, Thinh Hung Otmakhova, Yulia Verspoor, Karin Cohn, Trevor Baldwin, Timothy |
| author_facet | Truong, Thinh Hung Otmakhova, Yulia Verspoor, Karin Cohn, Trevor Baldwin, Timothy |
| contents | In this work, we measure the impact of affixal negation on modern English large language models (LLMs). In affixal negation, the negated meaning is expressed through a negative morpheme, which is potentially challenging for LLMs as their tokenizers are often not morphologically plausible. We conduct extensive experiments using LLMs with different subword tokenization methods, which lead to several insights on the interaction between tokenization performance and negation sensitivity. Despite some interesting mismatches between tokenization accuracy and negation detection performance, we show that models can, on the whole, reliably recognize the meaning of affixal negation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_02421 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Revisiting subword tokenization: A case study on affixal negation in large language models Truong, Thinh Hung Otmakhova, Yulia Verspoor, Karin Cohn, Trevor Baldwin, Timothy Computation and Language In this work, we measure the impact of affixal negation on modern English large language models (LLMs). In affixal negation, the negated meaning is expressed through a negative morpheme, which is potentially challenging for LLMs as their tokenizers are often not morphologically plausible. We conduct extensive experiments using LLMs with different subword tokenization methods, which lead to several insights on the interaction between tokenization performance and negation sensitivity. Despite some interesting mismatches between tokenization accuracy and negation detection performance, we show that models can, on the whole, reliably recognize the meaning of affixal negation. |
| title | Revisiting subword tokenization: A case study on affixal negation in large language models |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2404.02421 |