AAC Physics: Structure-Preserving Framework for Exact, Reversible Discrete-Time Convolutions

Published March 07, 2026 Version 1
Screened Endorsed AI Review Peer Review Accepted

Abstract

Memory-dependent discrete-time systems arise across physics, engineering, and machine learning—from viscoelasticity and control to differentiable programming. Standard recursive approaches accelerate convolution computation but suffer from drift, unstable backward passes, and numerical errors. (note: this paper was written using AI based on MY idea)

Loading PDF...

This may take a moment for large files

Comments

You must be logged in to comment

Login with ORCID

No comments yet. Be the first to comment!

Review Status

Stage 1

Awaiting Endorsement

Needs a Bronze+ ORCID scholar endorsement to advance.

Authors

Endorsements

No endorsements yet. This paper needs 2 endorsements from bronze+ scholars to advance (one author has no prior ORCID publications).

Endorse This Paper

You'll be asked to log in with ORCID.

Academic Categories

Algorithms

Formal Sciences > Computer Science > Theory of Computation > Algorithms

Machine Learning

Formal Sciences > Computer Science > Artificial Intelligence > Machine Learning

Stats

Versions 1
Comments 0
Authors 1