🐂 牛哥精选

Good-Enough LLM Obfuscation (GELO)

新方法GELO专为共享加速器场景设计，有效保护开源大模型的提示词隐私，对抗内存读取攻击。

arXiv:2603.05035v3 Announce Type: replace-cross Abstract: Large Language Models (LLMs) are increasingly served on shared accelerators where an adversa…