Formal Fields: A Framework to Automate Code Generation Across Domains

TL;DR

Formal Fields is a unified framework that enables cross-domain code generation by learning domain-specific primitives and structures, using reinforcement learning, to solve previously unsolved problems in a flexible, abstract manner.

Contribution

It introduces a domain-agnostic framework for code generation that leverages reinforcement learning and abstraction to solve diverse problems across domains.

Findings

Solved 22 previously unsolved ARC problems

Implemented as an open-source project

Uses Monte-Carlo Tree Search for code exploration

Abstract

Code generation, defined as automatically writing a piece of code to solve a given problem for which an evaluation function exists, is a classic hard AI problem. Its general form, writing code using a general language used by human programmers from scratch is thought to be impractical. Adding constraints to the code grammar, implementing domain specific concepts as primitives and providing examples for the algorithm to learn, makes it practical. Formal fields is a framework to do code generation across domains using the same algorithms and language structure. Its ultimate goal is not just solving different narrow problems, but providing necessary abstractions to integrate many working solutions as a single lifelong reasoning system. It provides a common grammar to define: a domain language, a problem and its evaluation. The framework learns from examples of code snippets about the structure of the domain language and searches completely new code snippets to solve unseen problems in the same field. Formal fields abstract the search algorithm away from the problem. The search algorithm is taken from existing reinforcement learning algorithms. In our implementation it is an apropos Monte-Carlo Tree Search (MCTS). We have implemented formal fields as a fully documented open source project applied to the Abstract Reasoning Challenge (ARC). The implementation found code snippets solving twenty two previously unsolved ARC problems.