Self-Explained Keywords Empower Large Language Models for Code Generation

TL;DR

This paper introduces SEK, a technique that enhances large language models' code generation by extracting and explaining key problem-specific keywords, leading to significant performance improvements across multiple benchmarks.

Contribution

SEK is a novel method that helps LLMs better understand low-frequency keywords in code problems by self-explaining and ranking them, improving code generation accuracy.

Findings

SEK improves Pass@1 from 85.4% to 93.3% on HumanEval.

SEK consistently enhances performance across five LLMs and three benchmarks.

SEK shifts LLMs' focus from low-frequency to high-frequency keywords.

Abstract

Large language models (LLMs) have achieved impressive performance in code generation. However, due to the long-tail distribution of LLMs' training data, low-frequency terms are typically underrepresented in the training process. Consequently, LLMs often misunderstand or overlook problem-specific, low-frequency keywords during code generation, compromising the accuracy of the generated code. To address this, we propose a novel technique named SEK(\textbf{S}elf-\textbf{E}xplained \textbf{K}eywords), which empowers an LLM for better code generation by extracting and explaining the key terms in the problem description with the LLM itself and ranking them based on frequency. Comprehensive experiments across three benchmarks, i.e., HumanEval(+), MBPP(+), and APPS, with five representative LLMs, show that SEK can significantly improve LLMs in code generation, yielding substantial and consistent gains. For instance, SEK improves the Pass@1 of DeepSeek-Coder-V2-Instruct from 85.4\% to 93.3\% on the Humaneval benchmark. Further analysis confirms that SEK enables the LLMs to shift their attention from low-frequency keywords to their corresponding high-frequency counterparts.