the Linear Representation and Superposition Hypotheses

Representational Capacity: Geometric Limits on Feature Representation in Transformer Language Models

Announce Type: new Abstract: Model dimension ($d_{model}$) is a fundamental hyperparameter in transformer language models, yet its role in setting the geometric limits of feature representation remains under-explored. Grounded in the Linear Representation and Superposition Hypotheses - which propose that models encode features as near-orthogonal directions in latent space - we develop a framework for estimating how many such directions a model can support. We first establish the embedding...