arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2303.04729cs.LGcs.CLcs.CR
收藏

窃取语言模型的解码算法

Stealing the Decoding Algorithms of Language Models

University of Massachusetts Amherst(马萨诸塞大学阿默斯特分校)

Ali Naseh, Kalpesh Krishna, Mohit Iyyer, Amir Houmansadr

AI总结:

本文首次提出一种低成本的API攻击方法,可从GPT-2、GPT-3等语言模型中窃取解码算法类型及超参数,费用低至数美元。

摘要:

现代语言模型(LM)生成文本的一个关键组成部分是解码算法的选择与调优。这些算法决定了如何根据LM生成的内部概率分布来生成文本。选择解码算法并调优其超参数的过程需要大量时间、人工和计算资源,并且还需要广泛的人类评估。因此,此类解码算法的身份和超参数对其所有者而言被认为极具价值。在这项工作中,我们首次证明,对LM具有典型API访问权限的对手能够以极低的金钱成本窃取其解码算法的类型和超参数。我们的攻击对文本生成API中常用的流行LM有效,包括GPT-2、GPT-3和GPT-Neo。我们证明了仅需几美元即可窃取此类信息的可行性,例如,针对GPT-3的四个版本分别仅需$0.8、$1、$4和$40。

英文摘要:

A key component of generating text from modern language models (LM) is the selection and tuning of decoding algorithms. These algorithms determine how to generate text from the internal probability distribution generated by the LM. The process of choosing a decoding algorithm and tuning its hyperparameters takes significant time, manual effort, and computation, and it also requires extensive human evaluation. Therefore, the identity and hyperparameters of such decoding algorithms are considered to be extremely valuable to their owners. In this work, we show, for the first time, that an adversary with typical API access to an LM can steal the type and hyperparameters of its decoding algorithms at very low monetary costs. Our attack is effective against popular LMs used in text generation APIs, including GPT-2, GPT-3 and GPT-Neo. We demonstrate the feasibility of stealing such information with only a few dollars, e.g., $\$0.8$, $\$1$, $\$4$, and $\$40$ for the four versions of GPT-3.

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。

↑