Relevance Transformer: Generating Concise Code Snippets with Relevance Feedback

TL;DR

This paper introduces the Relevance Transformer, a novel Transformer-based model for automatic code generation that incorporates pseudo-relevance feedback to improve output relevance and diversity, demonstrating superior performance on standard benchmarks.

Contribution

The paper proposes the Relevance Transformer, which integrates external knowledge via pseudo-relevance feedback into Transformer architectures for enhanced code generation.

Findings

Outperforms state-of-the-art methods on BLEU scores

Effectively incorporates pseudo-relevance feedback during inference

Shows potential of Transformer architectures for code generation

Abstract

Tools capable of automatic code generation have the potential to augment programmer's capabilities. While straightforward code retrieval is incorporated into many IDEs, an emerging area is explicit code generation. Code generation is currently approached as a Machine Translation task, with Recurrent Neural Network (RNN) based encoder-decoder architectures trained on code-description pairs. In this work we introduce and study modern Transformer architectures for this task. We further propose a new model called the Relevance Transformer that incorporates external knowledge using pseudo-relevance feedback. The Relevance Transformer biases the decoding process to be similar to existing retrieved code while enforcing diversity. We perform experiments on multiple standard benchmark datasets for code generation including Django, Hearthstone, and CoNaLa. The results show improvements over state-of-the-art methods based on BLEU evaluation. The Relevance Transformer model shows the potential of Transformer-based architectures for code generation and introduces a method of incorporating pseudo-relevance feedback during inference.