Detecting Machine-Generated Long-Form Content with Latent-Space Variables
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917795392913408 |
|---|---|
| author | Tian, Yufei Pan, Zeyu Peng, Nanyun |
| author_facet | Tian, Yufei Pan, Zeyu Peng, Nanyun |
| contents | The increasing capability of large language models (LLMs) to generate fluent long-form texts is presenting new challenges in distinguishing machine-generated outputs from human-written ones, which is crucial for ensuring authenticity and trustworthiness of expressions. Existing zero-shot detectors primarily focus on token-level distributions, which are vulnerable to real-world domain shifts, including different prompting and decoding strategies, and adversarial attacks. We propose a more robust method that incorporates abstract elements, such as event transitions, as key deciding factors to detect machine versus human texts by training a latent-space model on sequences of events or topics derived from human-written texts. In three different domains, machine-generated texts, which are originally inseparable from human texts on the token level, can be better distinguished with our latent-space model, leading to a 31% improvement over strong baselines such as DetectGPT. Our analysis further reveals that, unlike humans, modern LLMs like GPT-4 generate event triggers and their transitions differently, an inherent disparity that helps our method to robustly detect machine-generated texts. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_03856 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Detecting Machine-Generated Long-Form Content with Latent-Space Variables Tian, Yufei Pan, Zeyu Peng, Nanyun Computation and Language Machine Learning The increasing capability of large language models (LLMs) to generate fluent long-form texts is presenting new challenges in distinguishing machine-generated outputs from human-written ones, which is crucial for ensuring authenticity and trustworthiness of expressions. Existing zero-shot detectors primarily focus on token-level distributions, which are vulnerable to real-world domain shifts, including different prompting and decoding strategies, and adversarial attacks. We propose a more robust method that incorporates abstract elements, such as event transitions, as key deciding factors to detect machine versus human texts by training a latent-space model on sequences of events or topics derived from human-written texts. In three different domains, machine-generated texts, which are originally inseparable from human texts on the token level, can be better distinguished with our latent-space model, leading to a 31% improvement over strong baselines such as DetectGPT. Our analysis further reveals that, unlike humans, modern LLMs like GPT-4 generate event triggers and their transitions differently, an inherent disparity that helps our method to robustly detect machine-generated texts. |
| title | Detecting Machine-Generated Long-Form Content with Latent-Space Variables |
| topic | Computation and Language Machine Learning |
| url | https://arxiv.org/abs/2410.03856 |