Latent Predictor Networks for Code Generation

TL;DR

The paper introduces Latent Predictor Networks, a neural architecture for code generation conditioned on diverse inputs, demonstrating improved performance on new datasets from card games.

Contribution

It proposes a scalable neural network framework that marginalizes over conditioning contexts and generation granularities for effective code synthesis.

Findings

Outperforms strong benchmarks on new datasets from Magic the Gathering and Hearthstone.

Demonstrates the effectiveness of marginalizing multiple predictors in code generation.

Provides new datasets for structured natural language and code generation tasks.

Abstract

Many language generation tasks require the production of text conditioned on both structured and unstructured inputs. We present a novel neural network architecture which generates an output sequence conditioned on an arbitrary number of input functions. Crucially, our approach allows both the choice of conditioning context and the granularity of generation, for example characters or tokens, to be marginalised, thus permitting scalable and effective training. Using this framework, we address the problem of generating programming code from a mixed natural language and structured specification. We create two new data sets for this paradigm derived from the collectible trading card games Magic the Gathering and Hearthstone. On these, and a third preexisting corpus, we demonstrate that marginalising multiple predictors allows our model to outperform strong benchmarks.