全吠无咬：Transformer语言模型中的不规则维度模糊表示质量
All Bark and No Bite: Rogue Dimensions in Transformer Language Models Obscure Representational Quality
摘要：
相似性度量是理解语言模型如何表示和处理语言的重要工具。标准的表示相似性度量，如余弦相似性和欧氏距离，已成功地用于静态词嵌入模型，以了解如何在语义空间中的词集群。最近，这些措施已被应用到嵌入的上下文模型，如BERT和GPT-2。在这项工作中，我们提出了这样的措施的信息性的问题，语境化的语言模型。我们发现，少数的流氓维度，往往只有1-3，这些措施占主导地位。此外，我们发现主导相似性度量的维度与对模型行为很重要的维度之间存在显著的不匹配。我们表明，简单的后处理技术，如标准化能够纠正流氓尺寸，并揭示潜在的代表性质量。我们认为，占流氓尺寸是必不可少的任何相似性为基础的分析上下文语言模型。
英文摘要：
Similarity measures are a vital tool for understanding how language models represent and process language. Standard representational similarity measures such as cosine similarity and Euclidean distance have been successfully used in static word embedding models to understand how words cluster in semantic space. Recently, these measures have been applied to embeddings from contextualized models such as BERT and GPT-2. In this work, we call into question the informativity of such measures for contextualized language models. We find that a small number of rogue dimensions, often just 1-3, dominate these measures. Moreover, we find a striking mismatch between the dimensions that dominate similarity measures and those which are important to the behavior of the model. We show that simple postprocessing techniques such as standardization are able to correct for rogue dimensions and reveal underlying representational quality. We argue that accounting for rogue dimensions is essential for any similarity-based analysis of contextual language models.
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。
