Improved Tree Search for Automatic Program Synthesis

03/13/2023
by   Aran Carmon, et al.
0

In the task of automatic program synthesis, one obtains pairs of matching inputs and outputs and generates a computer program, in a particular domain-specific language (DSL), which given each sample input returns the matching output. A key element is being able to perform an efficient search in the space of valid programs. Here, we suggest a variant of MCTS that leads to state of the art results on two vastly different DSLs. The exploration method we propose includes multiple contributions: a modified visit count, a preprocessing procedure for the training dataset, and encoding the part of the program that was already executed.

READ FULL TEXT

page 6

page 7

page 9

research
06/19/2020

Neural Program Synthesis with a Differentiable Fixer

We present a new program synthesis approach that combines an encoder-dec...
research
10/04/2020

Optimal Neural Program Synthesis from Multimodal Specifications

Multimodal program synthesis, which leverages different types of user in...
research
12/02/2016

Summary - TerpreT: A Probabilistic Programming Language for Program Induction

We study machine learning formulations of inductive program synthesis; t...
research
05/26/2022

Tensor Program Optimization with Probabilistic Programs

Automatic optimization for tensor programs becomes increasingly importan...
research
09/12/2018

Automatic Program Synthesis of Long Programs with a Learned Garbage Collector

We consider the problem of generating automatic code given sample input-...
research
07/25/2023

Synthesis of Procedural Models for Deterministic Transition Systems

This paper introduces a general approach for synthesizing procedural mod...
research
06/22/2020

Information-theoretic User Interaction: Significant Inputs for Program Synthesis

Programming-by-example technologies are being deployed in industrial pro...

Please sign up or login with your details

Forgot password? Click here to reset