Process-Tensor Tomography of SGD: Measuring Non-Markovian Memory via Back-Flow of Distinguishability

Vasileios Sevetlidis*, George Pavlidis · 2026

Read Process-Tensor Tomography of SGD: Measuring Non-Markovian Memory via Back-Flow of Distinguishability by Vasileios Sevetlidis*, George Pavlidis (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-16563.txt.

Description

Vasileios Sevetlidis and George Pavlidis view SGD train as a process tensor. Their 2026 preprint deals with how nets train step by step. SGD is the main train rule for nets in wide use. A process tensor maps a chain of acts to a read out. Here acts span batch picks, tweaks and step parts of the rule.

That view lets them sum past pull in the train path. Past pull is read by how clear past states stay in time. Clear pull marks mean train has a long past tail. Such tails shape how nets learn and what they keep. The work gives a new lens on train acts.

Catalog record

CREATORVasileios Sevetlidis*, George Pavlidis
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE60 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-16563

Browse the full library · About the Tower of Babel