LLM Trojans that Increase Response Lengths and Cost

Zhengxing Li ,

Guangmingmei Yang ,

Nader Alfares ,

David J. Miller ,

George Kesidis

LLM Other LLM AI Safety / AI Ethics PIAAP MBDSF

如何管控大语言模型（LLM）使用过程中基于 Token 的费用，已成为一个日益突出的重要问题。部分前沿模型提供商对每个响应 Token 的收费高达每个提示 Token 的五倍；而另一些提供商则为提示 Token 提供预填充（prefilling）折扣。本文提出一种防御机制，用以应对一种新型数据投毒攻击：该攻击在模型中植入“木马”，当提示中出现特定触发器时，便会诱导模型生成更长的响应（从而增加用户的使用成本）。具体而言，我们展示了：针对经过指令微调的大语言模型，一种隐蔽性极强的木马可被语义上相似的各类指令所激活。

Management of the token-based costs of using LLMs has emerged as a significant problem. Some frontier-model providers charge five times per response token than per prompt token, while others offer prefilling discounts for prompt tokens. In this paper, we devise a defense against a new data-poisoning attack that plants a Trojan which increases the length of the response (and hence the cost to the user) when its trigger is present in the prompt. In particular, we demonstrate how a subtle Trojan may trigger on semantically similar instructions of an instruction fine-tuned LLM.