Saved in:
Bibliographic Details
Main Authors: Tang, Bo, Xie, Weiwei
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2603.19272
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908901765545984
author Tang, Bo
Xie, Weiwei
author_facet Tang, Bo
Xie, Weiwei
contents Differentiable Neural Computers (DNCs) were introduced as recurrent architectures equipped with an addressable external memory supporting differentiable read and write operations. Transformers, in contrast, are nominally feedforward architectures based on multi-head self-attention. In this work we give a formal derivation showing that a causal Transformer layer is exactly a stateless Differentiable Neural Computer (sDNC) where (1) the controller has no recurrent internal state, (2) the external memory is a write-once matrix of value vectors, (3) content-based addressing via keys implements attention, and (4) multi-head attention corresponds to multiple parallel read heads. We further extend this equivalence to cross-attention, showing that encoder-decoder Transformers are precisely sDNCs with distinct read-from and write-to memories. Our results provide a unified memory-centric interpretation of Transformers and contribute to the ongoing effort to place modern large language models in a principled computational framework.
format Preprint
id arxiv_https___arxiv_org_abs_2603_19272
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Transformers are Stateless Differentiable Neural Computers
Tang, Bo
Xie, Weiwei
Computation and Language
Artificial Intelligence
Computer Vision and Pattern Recognition
Machine Learning
Differentiable Neural Computers (DNCs) were introduced as recurrent architectures equipped with an addressable external memory supporting differentiable read and write operations. Transformers, in contrast, are nominally feedforward architectures based on multi-head self-attention. In this work we give a formal derivation showing that a causal Transformer layer is exactly a stateless Differentiable Neural Computer (sDNC) where (1) the controller has no recurrent internal state, (2) the external memory is a write-once matrix of value vectors, (3) content-based addressing via keys implements attention, and (4) multi-head attention corresponds to multiple parallel read heads. We further extend this equivalence to cross-attention, showing that encoder-decoder Transformers are precisely sDNCs with distinct read-from and write-to memories. Our results provide a unified memory-centric interpretation of Transformers and contribute to the ongoing effort to place modern large language models in a principled computational framework.
title Transformers are Stateless Differentiable Neural Computers
topic Computation and Language
Artificial Intelligence
Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2603.19272