The Gradient of Algebraic Model Counting

Jaron Maene, Luc De Raedt · 2025

Read The Gradient of Algebraic Model Counting by Jaron Maene, Luc De Raedt (2025) 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-2502-18406.txt.

Description

Jaron Maene and Luc De Raedt present a study from twenty twenty-five on how a unifying frame for logic based sums also covers learning. Algebraic model counting adds up the worth of logic forms with ring like math tools. It joins many inference tasks in one view. The pair ask if the same view can join learning methods too.

They show that gradients and backpropagation stretch to new rings in the same way. Traits of a ring, like when terms cancel or rank, can cut the memory cost of the backward pass. That yields fresh forms of top gradient methods for logic based chance models. Tests show clear speed gains over past routes. They also note why second order methods gain less here.

Catalog record

CREATORJaron Maene, Luc De Raedt
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE46 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2502-18406

Browse the full library · About the Tower of Babel