论文标题
用噪音进行探测：取消嵌入的翘曲和纬线
Probing with Noise: Unpicking the Warp and Weft of Embeddings
论文作者
论文摘要
提高我们对向量空间中信息的理解可以产生有价值的解释性见解。除了矢量维度外，我们认为矢量规范也有可能携带语言信息。我们开发了一种测试方法：探测框架的扩展，该探测框架允许对探测结果的相对内在解释。它依赖于引入噪声，即以随机基线和置信区间为基础编码的嵌入信息。我们将该方法应用于良好的探测任务，并找到证据，以证实英语手套和伯特嵌入中的单独信息容器的存在。我们的相关分析与实验发现一致，即不同编码器使用规范编码不同种类的信息：手套存储在矢量规范中的句法和句子长度信息，而Bert则使用它来编码上下文不一致。
Improving our understanding of how information is encoded in vector space can yield valuable interpretability insights. Alongside vector dimensions, we argue that it is possible for the vector norm to also carry linguistic information. We develop a method to test this: an extension of the probing framework which allows for relative intrinsic interpretations of probing results. It relies on introducing noise that ablates information encoded in embeddings, grounded in random baselines and confidence intervals. We apply the method to well-established probing tasks and find evidence that confirms the existence of separate information containers in English GloVe and BERT embeddings. Our correlation analysis aligns with the experimental findings that different encoders use the norm to encode different kinds of information: GloVe stores syntactic and sentence length information in the vector norm, while BERT uses it to encode contextual incongruity.
