CYCLE: Learning to Self-Refine the Code Generation

• URL: http://arxiv.org/abs/2403.18746v1
• Date: Wed, 27 Mar 2024 16:45:02 GMT
• Title: CYCLE: Learning to Self-Refine the Code Generation
• Authors: Yangruibo Ding, Marcus J. Min, Gail Kaiser, Baishakhi Ray,
• Abstract summary: We propose CYCLE framework, learning to self-refine the faulty generation according to the available feedback. We implement four variants of CYCLE with varied numbers of parameters across 350M, 1B, 2B, and 3B benchmarks. The results reveal that CYCLE successfully maintains, sometimes improves, the quality of one-time code generation, while significantly improving the self-refinement capability of code LMs.
• Score: 19.71833229434497
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Pre-trained code language models have achieved promising performance in code generation and improved the programming efficiency of human developers. However, their self-refinement capability is typically overlooked by the existing evaluations of code LMs, which focus only on the accuracy of the one-time prediction. For the cases when code LMs fail to implement the correct program, developers actually find it hard to debug and fix the faulty prediction since it is not written by the developers themselves. Unfortunately, our study reveals that code LMs cannot efficiently self-refine their faulty generations as well. In this paper, we propose CYCLE framework, learning to self-refine the faulty generation according to the available feedback, such as the execution results reported by the test suites. We evaluate CYCLE on three popular code generation benchmarks, HumanEval, MBPP, and APPS. The results reveal that CYCLE successfully maintains, sometimes improves, the quality of one-time code generation, while significantly improving the self-refinement capability of code LMs. We implement four variants of CYCLE with varied numbers of parameters across 350M, 1B, 2B, and 3B, and the experiments show that CYCLE consistently boosts the code generation performance, by up to 63.5%, across benchmarks and varied model sizes. We also notice that CYCLE outperforms code LMs that have 3$\times$ more parameters in self-refinement.

Related papers

• Prompting and Fine-tuning Large Language Models for Automated Code Review Comment Generation [5.6001617185032595]
  Large language models pretrained on both programming and natural language data tend to perform well in code-oriented tasks. We fine-tune open-source Large language models (LLM) in parameter-efficient, quantized low-rank fashion on consumer-grade hardware to improve review comment generation.
  arXiv  Detail & Related papers  (2024-11-15T12:01:38Z)
• No Man is an Island: Towards Fully Automatic Programming by Code Search, Code Generation and Program Repair [9.562123938545522]
  toolname can integrate various code search, generation, and repair tools, combining these three research areas together for the first time. We conduct preliminary experiments to demonstrate the potential of our framework, eg helping CodeLlama solve 267 programming problems with an improvement of 62.53%.
  arXiv  Detail & Related papers  (2024-09-05T06:24:29Z)
• Hotfixing Large Language Models for Code [8.243596444097506]
  Large Language Models for Code (LLM4Code) have become an integral part of developers', assisting with tasks such as code completion and generation. These models are found to exhibit undesired behaviors after their release, like generating buggy code. This paper mainly focuses on hotfixing LLM4Code to make them generate less buggy code and more fixed code.
  arXiv  Detail & Related papers  (2024-08-11T08:34:43Z)
• What's Wrong with Your Code Generated by Large Language Models? An Extensive Study [80.18342600996601]
  Large language models (LLMs) produce code that is shorter yet more complicated as compared to canonical solutions. We develop a taxonomy of bugs for incorrect codes that includes three categories and 12 sub-categories, and analyze the root cause for common bug types. We propose a novel training-free iterative method that introduces self-critique, enabling LLMs to critique and correct their generated code based on bug types and compiler feedback.
  arXiv  Detail & Related papers  (2024-07-08T17:27:17Z)
• Reasoning Runtime Behavior of a Program with LLM: How Far Are We? [25.451857140926943]
  Large language models for code (i.e., code LLMs) have shown strong code understanding and generation capabilities. Code reasoning is one of the most essential abilities of code LLMs. We propose a framework, namely REval, for evaluating code reasoning abilities and consistency of code LLMs with program execution.
  arXiv  Detail & Related papers  (2024-03-25T05:37:16Z)
• Assured LLM-Based Software Engineering [51.003878077888686]
  This paper is an outline of the content of the keynote by Mark Harman at the International Workshop on Interpretability, Robustness, and Benchmarking in Neural Software Engineering, Monday 15th April 2024, Lisbon, Portugal.
  arXiv  Detail & Related papers  (2024-02-06T20:38:46Z)
• Is Your Code Generated by ChatGPT Really Correct? Rigorous Evaluation of Large Language Models for Code Generation [20.45045253933097]
  We propose EvalPlus -- a code synthesis evaluation framework to rigorously benchmark the functional correctness of LLM-synthesized code. EvalPlus augments a given evaluation dataset with large amounts of test-cases newly produced by an automatic test input generator. We show that HumanEval+ is able to catch significant amounts of previously undetected wrong code.
  arXiv  Detail & Related papers  (2023-05-02T05:46:48Z)
• Teaching Large Language Models to Self-Debug [62.424077000154945]
  Large language models (LLMs) have achieved impressive performance on code generation. We propose Self- Debugging, which teaches a large language model to debug its predicted program via few-shot demonstrations.
  arXiv  Detail & Related papers  (2023-04-11T10:43:43Z)
• ReCode: Robustness Evaluation of Code Generation Models [90.10436771217243]
  We propose ReCode, a comprehensive robustness evaluation benchmark for code generation models. We customize over 30 transformations specifically for code on docstrings, function and variable names, code syntax, and code format. With human annotators, we verified that over 90% of the perturbed prompts do not alter the semantic meaning of the original prompt.
  arXiv  Detail & Related papers  (2022-12-20T14:11:31Z)
• Fault-Aware Neural Code Rankers [64.41888054066861]
  We propose fault-aware neural code rankers that can predict the correctness of a sampled program without executing it. Our fault-aware rankers can significantly increase the pass@1 accuracy of various code generation models.
  arXiv  Detail & Related papers  (2022-06-04T22:01:05Z)
• Measuring Coding Challenge Competence With APPS [54.22600767666257]
  We introduce APPS, a benchmark for code generation. Our benchmark includes 10,000 problems, which range from having simple one-line solutions to being substantial algorithmic challenges. Recent models such as GPT-Neo can pass approximately 15% of the test cases of introductory problems.
  arXiv  Detail & Related papers  (2021-05-20T17:58:42Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.