Top-Down Synthesis for Library Learning

TL;DR

This paper presents Stitch, a top-down, corpus-guided synthesis algorithm for library learning that is significantly faster and more memory-efficient than previous deductive methods, while maintaining high-quality abstractions.

Contribution

The paper introduces Stitch, a novel top-down synthesis approach for library learning that outperforms existing deductive algorithms in speed, memory usage, and scalability.

Findings

Stitch is 3-4 orders of magnitude faster than DreamCoder.

Stitch uses 2 orders of magnitude less memory.

Stitch scales to large corpora with hundreds of programs.

Abstract

This paper introduces corpus-guided top-down synthesis as a mechanism for synthesizing library functions that capture common functionality from a corpus of programs in a domain specific language (DSL). The algorithm builds abstractions directly from initial DSL primitives, using syntactic pattern matching of intermediate abstractions to intelligently prune the search space and guide the algorithm towards abstractions that maximally capture shared structures in the corpus. We present an implementation of the approach in a tool called Stitch and evaluate it against the state-of-the-art deductive library learning algorithm from DreamCoder. Our evaluation shows that Stitch is 3-4 orders of magnitude faster and uses 2 orders of magnitude less memory while maintaining comparable or better library quality (as measured by compressivity). We also demonstrate Stitch's scalability on corpora containing hundreds of complex programs that are intractable with prior deductive approaches and show empirically that it is robust to terminating the search procedure early -- further allowing it to scale to challenging datasets by means of early stopping.