arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2406.05532cs.LGcs.AI
收藏

深度状态空间模型的对抗鲁棒性探究

Exploring Adversarial Robustness of Deep State Space Models

Tsinghua University(清华大学);Shanghai Artificial Intelligence Laboratory(上海人工智能实验室);Harbin Institute of Technology(哈尔滨工业大学)

Biqing Qi, Yang Luo, Junqi Gao, Pengfei Li, Kai Tian, Zhiyuan Ma, Bowen Zhou

AI总结:

该研究探究深度状态空间模型的对抗鲁棒性,分析不同结构变体在对抗训练中的表现及误差机制,提出自适应缩放机制,在提升对抗训练效果的同时避免鲁棒过拟合问题。

摘要:

深度状态空间模型(Deep State Space Models, SSMs)已在众多任务场景中被证明有效,但在实际部署中会因对抗扰动(Adversarial Perturbations, APs)面临严峻的安全挑战。对抗训练(Adversarial Training, AT)是提升对抗鲁棒性(Adversarial Robustness, AR)的主流方法,已在多种传统DNN架构上得到验证,但其在提升SSMs对抗鲁棒性方面的效果仍不明确。尽管SSM组件的诸多改进(如集成注意力机制、扩展为数据依赖的SSM参数化)在标准训练(Standard Training, ST)设置下带来了显著收益,但它们在对抗训练中的潜在优势尚未被探究。\n为研究这一问题,我们对现有SSM结构变体结合对抗训练进行评估,以测试其对抗鲁棒性能。我们观察到纯SSM结构难以从对抗训练中获益,而与其他组件相比,集成注意力机制能让对抗训练中的SSM在鲁棒性和泛化性之间取得明显更优的权衡。不过,注意力的集成也会引发鲁棒过拟合(Robust Overfitting, RO)问题。\n为理解这些现象,我们从实证和理论层面分析了SSM在对抗扰动下的输出误差。我们发现固定参数化的SSM的输出误差界与其参数严格相关,这限制了它们的对抗训练收益;而输入依赖型SSM可能面临误差爆炸的问题。此外,我们证明注意力组件在训练过程中能有效缩放SSM的输出误差,使其能从对抗训练中获益更多,但代价是因模型复杂度较高而引入鲁棒过拟合。受此启发,我们提出了一种简单有效的自适应缩放(Adaptive Scaling, AdS)机制,能让对抗训练性能接近集成注意力的SSM,且不会引入鲁棒过拟合问题。我们的代码可在https://github.com/Biqing-Qi/Exploring-Adversarial-Robustness-of-Deep-State-Space-Models.git获取。

英文摘要:

Deep State Space Models (SSMs) have proven effective in numerous task scenarios but face significant security challenges due to Adversarial Perturbations (APs) in real-world deployments. Adversarial Training (AT) is a mainstream approach to enhancing Adversarial Robustness (AR) and has been validated on various traditional DNN architectures. However, its effectiveness in improving the AR of SSMs remains unclear. While many enhancements in SSM components, such as integrating Attention mechanisms and expanding to data-dependent SSM parameterizations, have brought significant gains in Standard Training (ST) settings, their potential benefits in AT remain unexplored. To investigate this, we evaluate existing structural variants of SSMs with AT to assess their AR performance. We observe that pure SSM structures struggle to benefit from AT, whereas incorporating Attention yields a markedly better trade-off between robustness and generalization for SSMs in AT compared to other components. Nonetheless, the integration of Attention also leads to Robust Overfitting (RO) issues. To understand these phenomena, we empirically and theoretically analyze the output error of SSMs under AP. We find that fixed-parameterized SSMs have output error bounds strictly related to their parameters, limiting their AT benefits, while input-dependent SSMs may face the problem of error explosion. Furthermore, we show that the Attention component effectively scales the output error of SSMs during training, enabling them to benefit more from AT, but at the cost of introducing RO due to its high model complexity. Inspired by this, we propose a simple and effective Adaptive Scaling (AdS) mechanism that brings AT performance close to Attention-integrated SSMs without introducing the issue of RO. Our code is available at https://github.com/Biqing-Qi/Exploring-Adversarial-Robustness-of-Deep-State-Space-Models.git.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑