Differentiation and Specialization of Attention Heads via the Refined Local Learning Coefficient
원문 초록
We introduce refined variants of the Local Learning Coefficient (LLC), a measure of model complexity grounded in singular learning theory, to study the development of internal structure in transformer language models during training. By applying these \textit{refined LLCs} (rLLCs) to individual components of a two-layer attention-only transformer, we gain novel insights into the progressive differentiation and specialization of attention heads. Our methodology reveals how attention heads differentiate into distinct functional roles over the course of training, analyzes the types of data these heads specialize to process, and discovers a previously unidentified multigram circuit. These findings demonstrate that rLLCs provide a principled, quantitative toolkit for \textit{developmental interpretability}, which aims to understand models through their evolution across the learning process. More broadly, this work takes a step towards establishing the correspondence between data distributional structure, geometric properties of the loss landscape, learning dynamics, and emergent computational structures in neural networks.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
Self-Supervised Evolution Operator Learning for High-Dimensional Dynamical Systems
auto-fpt: Automating Free Probability Theory Calculations for Machine Learning Theory
Double descent in quantum machine learning
Gradient Equilibrium in Online Learning: Theory and Applications
NLPrompt: Noise-Label Prompt Learning for Vision-Language Models
Federated Learning from Vision-Language Foundation Models: Theoretical Analysis and Method
토론
이 논문과 연결된 포럼 글이 아직 없습니다.