Grokking Modular Polynomials

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Doshi, Darshil, He, Tianyu, Das, Aritra, Gromov, Andrey
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866911906422325248
author Doshi, Darshil
He, Tianyu
Das, Aritra
Gromov, Andrey
author_facet Doshi, Darshil
He, Tianyu
Das, Aritra
Gromov, Andrey
contents Neural networks readily learn a subset of the modular arithmetic tasks, while failing to generalize on the rest. This limitation remains unmoved by the choice of architecture and training strategies. On the other hand, an analytical solution for the weights of Multi-layer Perceptron (MLP) networks that generalize on the modular addition task is known in the literature. In this work, we (i) extend the class of analytical solutions to include modular multiplication as well as modular addition with many terms. Additionally, we show that real networks trained on these datasets learn similar solutions upon generalization (grokking). (ii) We combine these "expert" solutions to construct networks that generalize on arbitrary modular polynomials. (iii) We hypothesize a classification of modular polynomials into learnable and non-learnable via neural networks training; and provide experimental evidence supporting our claims.
format Preprint
id arxiv_https___arxiv_org_abs_2406_03495
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Grokking Modular Polynomials
Doshi, Darshil
He, Tianyu
Das, Aritra
Gromov, Andrey
Machine Learning
Disordered Systems and Neural Networks
High Energy Physics - Theory
Number Theory
Neural networks readily learn a subset of the modular arithmetic tasks, while failing to generalize on the rest. This limitation remains unmoved by the choice of architecture and training strategies. On the other hand, an analytical solution for the weights of Multi-layer Perceptron (MLP) networks that generalize on the modular addition task is known in the literature. In this work, we (i) extend the class of analytical solutions to include modular multiplication as well as modular addition with many terms. Additionally, we show that real networks trained on these datasets learn similar solutions upon generalization (grokking). (ii) We combine these "expert" solutions to construct networks that generalize on arbitrary modular polynomials. (iii) We hypothesize a classification of modular polynomials into learnable and non-learnable via neural networks training; and provide experimental evidence supporting our claims.
title Grokking Modular Polynomials
topic Machine Learning
Disordered Systems and Neural Networks
High Energy Physics - Theory
Number Theory
url https://arxiv.org/abs/2406.03495