Exploring the Limitations of Detecting Machine-Generated Text
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_ | 1866909424991338496 |
|---|---|
| author | Doughman, Jad Afzal, Osama Mohammed Toyin, Hawau Olamide Shehata, Shady Nakov, Preslav Talat, Zeerak |
| author_facet | Doughman, Jad Afzal, Osama Mohammed Toyin, Hawau Olamide Shehata, Shady Nakov, Preslav Talat, Zeerak |
| contents | Recent improvements in the quality of the generations by large language models have spurred research into identifying machine-generated text. Such work often presents high-performing detectors. However, humans and machines can produce text in different styles and domains, yet the performance impact of such on machine generated text detection systems remains unclear. In this paper, we audit the classification performance for detecting machine-generated text by evaluating on texts with varying writing styles. We find that classifiers are highly sensitive to stylistic changes and differences in text complexity, and in some cases degrade entirely to random classifiers. We further find that detection systems are particularly susceptible to misclassify easy-to-read texts while they have high performance for complex texts, leading to concerns about the reliability of detection systems. We recommend that future work attends to stylistic factors and reading difficulty levels of human-written and machine-generated text. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_11073 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Exploring the Limitations of Detecting Machine-Generated Text Doughman, Jad Afzal, Osama Mohammed Toyin, Hawau Olamide Shehata, Shady Nakov, Preslav Talat, Zeerak Computation and Language Recent improvements in the quality of the generations by large language models have spurred research into identifying machine-generated text. Such work often presents high-performing detectors. However, humans and machines can produce text in different styles and domains, yet the performance impact of such on machine generated text detection systems remains unclear. In this paper, we audit the classification performance for detecting machine-generated text by evaluating on texts with varying writing styles. We find that classifiers are highly sensitive to stylistic changes and differences in text complexity, and in some cases degrade entirely to random classifiers. We further find that detection systems are particularly susceptible to misclassify easy-to-read texts while they have high performance for complex texts, leading to concerns about the reliability of detection systems. We recommend that future work attends to stylistic factors and reading difficulty levels of human-written and machine-generated text. |
| title | Exploring the Limitations of Detecting Machine-Generated Text |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2406.11073 |