arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2502.07586cs.CLcs.AI
收藏

我们无法用现有词汇理解人工智能

We Can't Understand AI Using our Existing Vocabulary

Google DeepMind(谷歌DeepMind)

John Hewitt, Robert Geirhos, Been Kim

AI总结:

本立场论文指出现有人类词汇无法理解AI,主张创造新词构建人机共享语言以解决可解释性的沟通问题,通过“长度”和“多样性”新词验证其能控制LLM响应并促进人机概念交互。

摘要:

本立场论文主张,为理解人工智能,我们不能依赖现有的人类词汇,而应努力创造新词:即代表我们想教给机器的精确人类概念,或我们需要学习的机器概念的新词汇。我们从人类与机器拥有不同概念这一前提出发,这意味着可解释性可被视为一个沟通问题:人类必须能够引用和控制机器概念,并向机器传达人类概念。我们认为,通过创造新词构建人机共享语言可解决此沟通问题。成功的新词需达到有用的抽象程度:既不过于详细(以便在多种语境中复用),也不过于宽泛(以便传达精确信息)。作为概念验证,我们展示了“长度新词”如何实现对LLM响应长度的控制,而“多样性新词”则能让模型生成更多样化的响应。综上,我们认为无法用现有词汇理解人工智能,通过新词扩展词汇可创造更好地控制和理解机器的机会。

英文摘要:

This position paper argues that, in order to understand AI, we cannot rely on our existing vocabulary of human words. Instead, we should strive to develop neologisms: new words that represent precise human concepts that we want to teach machines, or machine concepts that we need to learn. We start from the premise that humans and machines have differing concepts. This means interpretability can be framed as a communication problem: humans must be able to reference and control machine concepts, and communicate human concepts to machines. Creating a shared human-machine language through developing neologisms, we believe, could solve this communication problem. Successful neologisms achieve a useful amount of abstraction: not too detailed, so they're reusable in many contexts, and not too high-level, so they convey precise information. As a proof of concept, we demonstrate how a "length neologism" enables controlling LLM response length, while a "diversity neologism" allows sampling more variable responses. Taken together, we argue that we cannot understand AI using our existing vocabulary, and expanding it through neologisms creates opportunities for both controlling and understanding machines better.

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。

↑