Temperature Matters: Enhancing Watermark Robustness Against Paraphrasing Attacks
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866915363775578112 |
|---|---|
| author | Idrissi, Badr Youbi Millunzi, Monica Sorrenti, Amelia Baraldi, Lorenzo Dementieva, Daryna |
| author_facet | Idrissi, Badr Youbi Millunzi, Monica Sorrenti, Amelia Baraldi, Lorenzo Dementieva, Daryna |
| contents | In the present-day scenario, Large Language Models (LLMs) are establishing their presence as powerful instruments permeating various sectors of society. While their utility offers valuable support to individuals, there are multiple concerns over potential misuse. Consequently, some academic endeavors have sought to introduce watermarking techniques, characterized by the inclusion of markers within machine-generated text, to facilitate algorithmic identification. This research project is focused on the development of a novel methodology for the detection of synthetic text, with the overarching goal of ensuring the ethical application of LLMs in AI-driven text generation. The investigation commences with replicating findings from a previous baseline study, thereby underscoring its susceptibility to variations in the underlying generation model. Subsequently, we propose an innovative watermarking approach and subject it to rigorous evaluation, employing paraphrased generated text to asses its robustness. Experimental results highlight the robustness of our proposal compared to the~\cite{aarson} watermarking method. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_22623 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Temperature Matters: Enhancing Watermark Robustness Against Paraphrasing Attacks Idrissi, Badr Youbi Millunzi, Monica Sorrenti, Amelia Baraldi, Lorenzo Dementieva, Daryna Computation and Language Artificial Intelligence In the present-day scenario, Large Language Models (LLMs) are establishing their presence as powerful instruments permeating various sectors of society. While their utility offers valuable support to individuals, there are multiple concerns over potential misuse. Consequently, some academic endeavors have sought to introduce watermarking techniques, characterized by the inclusion of markers within machine-generated text, to facilitate algorithmic identification. This research project is focused on the development of a novel methodology for the detection of synthetic text, with the overarching goal of ensuring the ethical application of LLMs in AI-driven text generation. The investigation commences with replicating findings from a previous baseline study, thereby underscoring its susceptibility to variations in the underlying generation model. Subsequently, we propose an innovative watermarking approach and subject it to rigorous evaluation, employing paraphrased generated text to asses its robustness. Experimental results highlight the robustness of our proposal compared to the~\cite{aarson} watermarking method. |
| title | Temperature Matters: Enhancing Watermark Robustness Against Paraphrasing Attacks |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2506.22623 |