CoSPED: Consistent Soft Prompt Targeted Data Extraction and Defense
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909997368082432 |
|---|---|
| author | Yang, Zhuochen Fok, Kar Wai Thing, Vrizlynn L. L. |
| author_facet | Yang, Zhuochen Fok, Kar Wai Thing, Vrizlynn L. L. |
| contents | Large language models have gained widespread attention recently, but their potential security vulnerabilities, especially privacy leakage, are also becoming apparent. To test and evaluate for data extraction risks in LLM, we proposed CoSPED, short for Consistent Soft Prompt targeted data Extraction and Defense. We introduce several innovative components, including Dynamic Loss, Additive Loss, Common Loss, and Self Consistency Decoding Strategy, and tested to enhance the consistency of the soft prompt tuning process. Through extensive experimentation with various combinations, we achieved an extraction rate of 65.2% at a 50-token prefix comparison. Our comparisons of CoSPED with other reference works confirm our superior extraction rates. We evaluate CoSPED on more scenarios, achieving Pythia model extraction rate of 51.7% and introducing cross-model comparison. Finally, we explore defense through Rank-One Model Editing and achieve a reduction in the extraction rate to 1.6%, which proves that our analysis of extraction mechanisms can directly inform effective mitigation strategies against soft prompt-based attacks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_11137 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | CoSPED: Consistent Soft Prompt Targeted Data Extraction and Defense Yang, Zhuochen Fok, Kar Wai Thing, Vrizlynn L. L. Cryptography and Security Large language models have gained widespread attention recently, but their potential security vulnerabilities, especially privacy leakage, are also becoming apparent. To test and evaluate for data extraction risks in LLM, we proposed CoSPED, short for Consistent Soft Prompt targeted data Extraction and Defense. We introduce several innovative components, including Dynamic Loss, Additive Loss, Common Loss, and Self Consistency Decoding Strategy, and tested to enhance the consistency of the soft prompt tuning process. Through extensive experimentation with various combinations, we achieved an extraction rate of 65.2% at a 50-token prefix comparison. Our comparisons of CoSPED with other reference works confirm our superior extraction rates. We evaluate CoSPED on more scenarios, achieving Pythia model extraction rate of 51.7% and introducing cross-model comparison. Finally, we explore defense through Rank-One Model Editing and achieve a reduction in the extraction rate to 1.6%, which proves that our analysis of extraction mechanisms can directly inform effective mitigation strategies against soft prompt-based attacks. |
| title | CoSPED: Consistent Soft Prompt Targeted Data Extraction and Defense |
| topic | Cryptography and Security |
| url | https://arxiv.org/abs/2510.11137 |