Saved in:
Bibliographic Details
Main Authors: Xia, Eric, Kalita, Jugal
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2507.14640
Tags: Add Tag
No Tags, Be the first to tag this record!
Table of Contents:
  • A two-part affine approximation has been found to be a good approximation for transformer computations over certain subject object relations. Adapting the Bigger Analogy Test Set, we show that the linear transformation Ws, where s is a middle layer representation of a subject token and W is derived from model derivatives, is also able to accurately reproduce final object states for many relations. This linear technique is able to achieve 90% faithfulness on morphological relations, and we show similar findings multi-lingually and across models. Our findings indicate that some conceptual relationships in language models, such as morphology, are readily interpretable from latent space, and are sparsely encoded by cross-layer linear transformations.