Saved in:
Bibliographic Details
Main Authors: Guo, Zhenyu, Chen, Wenguang
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2501.00823
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913636135469056
author Guo, Zhenyu
Chen, Wenguang
author_facet Guo, Zhenyu
Chen, Wenguang
contents Transformers have achieved remarkable success across diverse domains, but their monolithic architecture presents challenges in interpretability, adaptability, and scalability. This paper introduces a novel modular Transformer architecture that explicitly decouples knowledge and reasoning through a generalized cross-attention mechanism to a globally shared knowledge base with layer-specific transformations, specifically designed for effective knowledge retrieval. Critically, we provide a rigorous mathematical derivation demonstrating that the Feed-Forward Network (FFN) in a standard Transformer is a specialized case (a closure) of this generalized cross-attention, revealing its role in implicit knowledge retrieval and validating our design. This theoretical framework provides a new lens for understanding FFNs and lays the foundation for future research exploring enhanced interpretability, adaptability, and scalability, enabling richer interplay with external knowledge bases and other systems.
format Preprint
id arxiv_https___arxiv_org_abs_2501_00823
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Decoupling Knowledge and Reasoning in Transformers: A Modular Architecture with Generalized Cross-Attention
Guo, Zhenyu
Chen, Wenguang
Machine Learning
Artificial Intelligence
Computation and Language
Transformers have achieved remarkable success across diverse domains, but their monolithic architecture presents challenges in interpretability, adaptability, and scalability. This paper introduces a novel modular Transformer architecture that explicitly decouples knowledge and reasoning through a generalized cross-attention mechanism to a globally shared knowledge base with layer-specific transformations, specifically designed for effective knowledge retrieval. Critically, we provide a rigorous mathematical derivation demonstrating that the Feed-Forward Network (FFN) in a standard Transformer is a specialized case (a closure) of this generalized cross-attention, revealing its role in implicit knowledge retrieval and validating our design. This theoretical framework provides a new lens for understanding FFNs and lays the foundation for future research exploring enhanced interpretability, adaptability, and scalability, enabling richer interplay with external knowledge bases and other systems.
title Decoupling Knowledge and Reasoning in Transformers: A Modular Architecture with Generalized Cross-Attention
topic Machine Learning
Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2501.00823