LoRA通过随机投影实现天然差分隐私，降低隐私泄露风险。
LoRA Provides Differential Privacy by Design via Random Sketching
- 将LoRA的低秩适配机制等价为带噪声梯度的微调。
- 适配秩越低、批大小越大，隐私保护水平越高。
- 适用于关注隐私安全的模型微调场景。
低秩适配（LoRA）被提出用于降低预训练语言模型微调的计算与内存开销。LoRA在预训练模型的部分参数中引入可训练的低秩矩阵（称为适配器）。本文从理论上证明，LoRA的低秩适配机制等价于使用带有噪声梯度的适配器微调，其噪声方差随适配秩 $r$ 的减小而增加。基于此理解，我们证明当适配矩阵 $A_\ ext{ℓ}$ 冻结时，LoRA具有内在的差分隐私。我们进一步揭示适配秩、批大小等因素影响其可保证的隐私水平。研究结果为理解LoRA的隐私鲁棒性提供了理论依据。
原文摘要 · Abstract (English)
Low-rank adaptation of language models has been proposed to reduce the computational and memory overhead of fine-tuning pre-trained language models. LoRA incorporates trainable low-rank matrices into some parameters of the pre-trained model, called adapters. In this work, we show theoretically that the low-rank adaptation mechanism of LoRA is equivalent to fine-tuning adapters with noisy batch gradients, with the noise variance being a decreasing function of adaptation rank ($r$). Motivated by this understanding, we prove inherent differential privacy for LoRA when adaptation matrices $A_\ell$ are frozen. We show that various factors, e.g., the adaptation rank and batch size, affect the guaranteed privacy level. Our findings provide useful insights into LoRA and uncovers the reason behind the robustness of models fine-tuned with LoRA to privacy attacks.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。