arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2209.00445cs.CLcs.AIcs.LG
收藏

通过概念化解读嵌入空间

Interpreting Embedding Spaces by Conceptualization

Technion – Israel Institute of Technology(以色列理工学院)

Adi Simhi, Shaul Markovitch

AI总结:

针对大语言模型生成的嵌入空间人类难以理解的问题,提出将潜在嵌入空间转换为可理解概念空间的方法,设计动态按需粒度生成算法与新评估方式,可用于模型语义对比、LLM层追踪等任务。

摘要:

文本计算解读的主要方法之一是将文本映射到某个嵌入空间中的向量,这类向量随后可用于多种文本处理任务。近年来,大多数嵌入空间都是大型语言模型(LLM)训练的产物。这类表示方式的一个主要缺点是人类无法理解其含义。理解嵌入空间对多项重要需求至关重要,包括调试嵌入方法并将其与替代方案对比,以及检测模型中隐藏的偏见。本文提出一种理解嵌入的新方法,将潜在嵌入空间转换为可理解的概念空间。我们提出一种算法,用于生成具有动态按需粒度的概念空间。我们设计了一种新的评估方法,使用人类评分者或基于LLM的评分者,证明概念化向量确实能表征原始潜在向量的语义。我们展示了该方法在多种任务中的应用,包括对比替代模型的语义、追踪LLM的各层。代码可在以下网址获取:https://github.com/adiSimhi/Interpreting-Embedding-Spaces-by-Conceptualization。

英文摘要:

One of the main methods for computational interpretation of a text is mapping it into a vector in some embedding space. Such vectors can then be used for a variety of textual processing tasks. Recently, most embedding spaces are a product of training large language models (LLMs). One major drawback of this type of representation is their incomprehensibility to humans. Understanding the embedding space is crucial for several important needs, including the need to debug the embedding method and compare it to alternatives, and the need to detect biases hidden in the model. In this paper, we present a novel method of understanding embeddings by transforming a latent embedding space into a comprehensible conceptual space. We present an algorithm for deriving a conceptual space with dynamic on-demand granularity. We devise a new evaluation method, using either human rater or LLM-based raters, to show that the conceptualized vectors indeed represent the semantics of the original latent ones. We show the use of our method for various tasks, including comparing the semantics of alternative models and tracing the layers of the LLM. The code is available online https://github.com/adiSimhi/Interpreting-Embedding-Spaces-by-Conceptualization.

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。

↑