EcoSearch: A Constant-Delay Best-First Search Algorithm for Program Synthesis

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Matricon, Théo, Fijalkow, Nathanaël, Lagarde, Guillaume
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916538976567296
author Matricon, Théo
Fijalkow, Nathanaël
Lagarde, Guillaume
author_facet Matricon, Théo
Fijalkow, Nathanaël
Lagarde, Guillaume
contents Many approaches to program synthesis perform a combinatorial search within a large space of programs to find one that satisfies a given specification. To tame the search space blowup, previous works introduced probabilistic and neural approaches to guide this combinatorial search by inducing heuristic cost functions. Best-first search algorithms ensure to search in the exact order induced by the cost function, significantly reducing the portion of the program space to be explored. We present a new best-first search algorithm called EcoSearch, which is the first constant-delay algorithm for pre-generation cost function: the amount of compute required between outputting two programs is constant, and in particular does not increase over time. This key property yields important speedups: we observe that EcoSearch outperforms its predecessors on two classic domains.
format Preprint
id arxiv_https___arxiv_org_abs_2412_17330
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle EcoSearch: A Constant-Delay Best-First Search Algorithm for Program Synthesis
Matricon, Théo
Fijalkow, Nathanaël
Lagarde, Guillaume
Machine Learning
Artificial Intelligence
Programming Languages
Many approaches to program synthesis perform a combinatorial search within a large space of programs to find one that satisfies a given specification. To tame the search space blowup, previous works introduced probabilistic and neural approaches to guide this combinatorial search by inducing heuristic cost functions. Best-first search algorithms ensure to search in the exact order induced by the cost function, significantly reducing the portion of the program space to be explored. We present a new best-first search algorithm called EcoSearch, which is the first constant-delay algorithm for pre-generation cost function: the amount of compute required between outputting two programs is constant, and in particular does not increase over time. This key property yields important speedups: we observe that EcoSearch outperforms its predecessors on two classic domains.
title EcoSearch: A Constant-Delay Best-First Search Algorithm for Program Synthesis
topic Machine Learning
Artificial Intelligence
Programming Languages
url https://arxiv.org/abs/2412.17330