In the realm of deep learning, understanding the latent space of language models (LMs) like transformers is crucial for refining their performance and interpretability. However, existing analyses often fall short in providing absolute and model-centric insights into LM semantics, and neglect essential aspects of LM adaption. In response, we introduce a pioneering method called vocabulary-defined semantics, which establishes a fixed reference frame within the LM latent space, ensuring absolute semantic analysis grounded in LM vocabulary. Our approach transcends prior relative analyses, leveraging LM vocabulary for model-centric insights. Furthermore, we propose a novel technique to compute logits, emphasizing differentiability and local isotropy, and introduce a neural clustering module for semantically calibrating data representations during LM adaptation. Through extensive experiments across diverse text understanding datasets, our approach surpasses state-of-the-art methods of retrieval-augmented generation and parameters-efficient finetuning, showcasing its efficacy and broad applicability. Our findings not only shed light on LM mechanics but also offer practical solutions for enhancing LM performance and interpretability.

翻译：在深度学习领域，理解语言模型（如Transformer）的潜空间对于提升其性能与可解释性至关重要。然而，现有分析往往在提供关于语言模型语义的绝对性与模型中心性洞察方面存在不足，且忽略了语言模型适配的关键方面。为此，我们提出一种开创性方法——词汇定义语义，该方法在语言模型潜空间内建立固定参考框架，确保基于语言模型词汇的绝对语义分析。我们的方法超越了先前的相对分析，利用语言模型词汇获取模型中心的洞见。此外，我们提出一种新的对数几率计算技术，强调可微性与局部各向同性，并引入神经聚类模块以在语言模型适配过程中对数据表示进行语义校准。通过在多种文本理解数据集上的广泛实验，我们的方法在检索增强生成与参数高效微调方面超越了现有最先进技术，展示了其有效性与广泛适用性。我们的发现不仅揭示了语言模型的运作机制，还为提升语言模型性能与可解释性提供了实用解决方案。