Mitigating Gradient Inversion Risks in Language Models via Token Obfuscation

作者：Xinguo Feng, Zhongkui Ma, Zihan Wang, Alsharif Abuadbba, Guangdong Bai · 年份：2026 · DOI：10.1145/3779208.3785389 · 被引用次数：1 · 研究领域：Topic Modeling、Adversarial Robustness in Machine Learning、Generative Adversarial Networks and Image Synthesis

Training and fine-tuning large-scale language models largely benefit from collaborative learning, but the approach has been proven vulnerable to gradient inversion attacks (GIAs), which allow adversaries to reconstruct private training data from shared gradients. Existing defenses mainly employ gradient perturbation techniques, e.g., noise injection or gradient pruning, to disrupt GIAs' direct mapping from gradient space to token space. However, these methods often fall short due to the retention of semantics similarity across gradient, embedding, and token spaces. Attackers can map proximate gradients into similar embeddings, and subsequently correspond them to tokens with similar semantics.