Uncovering Hidden Intentions: Exploring Prompt Recovery for Deeper Insights into Generated Texts
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_ | 1866909229548306432 |
|---|---|
| author | Give, Louis Zaoral, Timo Bruno, Maria Antonietta |
| author_facet | Give, Louis Zaoral, Timo Bruno, Maria Antonietta |
| contents | Today, the detection of AI-generated content is receiving more and more attention. Our idea is to go beyond detection and try to recover the prompt used to generate a text. This paper, to the best of our knowledge, introduces the first investigation in this particular domain without a closed set of tasks. Our goal is to study if this approach is promising. We experiment with zero-shot and few-shot in-context learning but also with LoRA fine-tuning. After that, we evaluate the benefits of using a semi-synthetic dataset. For this first study, we limit ourselves to text generated by a single model. The results show that it is possible to recover the original prompt with a reasonable degree of accuracy. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_15871 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Uncovering Hidden Intentions: Exploring Prompt Recovery for Deeper Insights into Generated Texts Give, Louis Zaoral, Timo Bruno, Maria Antonietta Computation and Language Artificial Intelligence Today, the detection of AI-generated content is receiving more and more attention. Our idea is to go beyond detection and try to recover the prompt used to generate a text. This paper, to the best of our knowledge, introduces the first investigation in this particular domain without a closed set of tasks. Our goal is to study if this approach is promising. We experiment with zero-shot and few-shot in-context learning but also with LoRA fine-tuning. After that, we evaluate the benefits of using a semi-synthetic dataset. For this first study, we limit ourselves to text generated by a single model. The results show that it is possible to recover the original prompt with a reasonable degree of accuracy. |
| title | Uncovering Hidden Intentions: Exploring Prompt Recovery for Deeper Insights into Generated Texts |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2406.15871 |