ExtendAttack: Attacking Servers of LRMs via Extending Reasoning
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866914169067929600 |
|---|---|
| author | Zhu, Zhenhao Liu, Yue Xu, Zhiwei Ma, Yingwei Gao, Hongcheng Chen, Nuo Guo, Yanpei Qu, Wenjie Xu, Huiying Kang, Zifeng Zhu, Xinzhong Zhang, Jiaheng |
| author_facet | Zhu, Zhenhao Liu, Yue Xu, Zhiwei Ma, Yingwei Gao, Hongcheng Chen, Nuo Guo, Yanpei Qu, Wenjie Xu, Huiying Kang, Zifeng Zhu, Xinzhong Zhang, Jiaheng |
| contents | Large Reasoning Models (LRMs) have demonstrated promising performance in complex tasks. However, the resource-consuming reasoning processes may be exploited by attackers to maliciously occupy the resources of the servers, leading to a crash, like the DDoS attack in cyber. To this end, we propose a novel attack method on LRMs termed ExtendAttack to maliciously occupy the resources of servers by stealthily extending the reasoning processes of LRMs. Concretely, we systematically obfuscate characters within a benign prompt, transforming them into a complex, poly-base ASCII representation. This compels the model to perform a series of computationally intensive decoding sub-tasks that are deeply embedded within the semantic structure of the query itself. Extensive experiments demonstrate the effectiveness of our proposed ExtendAttack. Remarkably, it significantly increases response length and latency, with the former increasing by over 2.7 times for the o3 model on the HumanEval benchmark. Besides, it preserves the original meaning of the query and achieves comparable answer accuracy, showing the stealthiness. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_13737 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | ExtendAttack: Attacking Servers of LRMs via Extending Reasoning Zhu, Zhenhao Liu, Yue Xu, Zhiwei Ma, Yingwei Gao, Hongcheng Chen, Nuo Guo, Yanpei Qu, Wenjie Xu, Huiying Kang, Zifeng Zhu, Xinzhong Zhang, Jiaheng Cryptography and Security Large Reasoning Models (LRMs) have demonstrated promising performance in complex tasks. However, the resource-consuming reasoning processes may be exploited by attackers to maliciously occupy the resources of the servers, leading to a crash, like the DDoS attack in cyber. To this end, we propose a novel attack method on LRMs termed ExtendAttack to maliciously occupy the resources of servers by stealthily extending the reasoning processes of LRMs. Concretely, we systematically obfuscate characters within a benign prompt, transforming them into a complex, poly-base ASCII representation. This compels the model to perform a series of computationally intensive decoding sub-tasks that are deeply embedded within the semantic structure of the query itself. Extensive experiments demonstrate the effectiveness of our proposed ExtendAttack. Remarkably, it significantly increases response length and latency, with the former increasing by over 2.7 times for the o3 model on the HumanEval benchmark. Besides, it preserves the original meaning of the query and achieves comparable answer accuracy, showing the stealthiness. |
| title | ExtendAttack: Attacking Servers of LRMs via Extending Reasoning |
| topic | Cryptography and Security |
| url | https://arxiv.org/abs/2506.13737 |