Program Repair with Minimal Edits Using CodeT5
Not on the map. Why?
Abstract (arXiv)
Programmers often struggle to identify and fix bugs in their programs. In recent years, many language models (LMs) have been proposed to fix erroneous programs and support error recovery. However, the LMs tend to generate solutions that differ from the original input programs. This leads to potential comprehension difficulties for users. In this paper, we propose an approach to suggest a correct program with minimal repair edits using CodeT5. We fine-tune a pre-trained CodeT5 on code pairs of wrong and correct programs and evaluate its performance with several baseline models. The experimental results show that the fine-tuned CodeT5 achieves a pass@100 of 91.95% and an average edit distance of the most similar correct program of 6.84, which indicates that at least one correct program can be suggested by generating 100 candidate programs. We demonstrate the effectiveness of LMs in suggesting program repair with minimal edits for solving introductory programming problems.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 32 of its 53 references that are in the map. First posted on arXiv 26 Sep 2023.
Builds on
32 of 53 references.
References · 32
- 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
- 2020Language Models are Few-Shot Learnerscited by 32,199
- 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
- 2019Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformercited by 14,491
- 2021Evaluating Large Language Models Trained on Codecited by 8,053
- 2023Self-Refine: Iterative Refinement with Self-Feedbackcited by 3,317
- 2021ViViT: A Video Vision Transformercited by 1,610
- 2019fairseq: A Fast, Extensible Toolkit for Sequence Modelingcited by 1,486
- 2022Competition-Level Code Generation with AlphaCodecited by 1,427
- 2020CodeBERT: A Pre-Trained Model for Programming and Natural Languagescited by 1,286
- 2021CodeT5: Identifier-aware Unified Pre-trained Encoder-Decoder Models for Code Understanding and Generationcited by 890
- 2022CodeGen: An Open Large Language Model for Code with Multi-Turn Program Synthesiscited by 876
20 more
- 2023StarCoder: may the source be with you!cited by 790
- 2021CodeXGLUE: A Machine Learning Benchmark Dataset for Code Understanding and Generationcited by 732
- 2023WizardCoder: Empowering Code Large Language Models with Evol-Instructcited by 728
- 2022InCoder: A Generative Model for Code Infilling and Synthesiscited by 492
- 2020CodeBLEU: a Method for Automatic Evaluation of Code Synthesiscited by 468
- 2023CodeGeeX: A Pre-Trained Model for Code Generation with Multilingual Benchmarking on HumanEval-Xcited by 294
- 2020Unsupervised Translation of Programming Languagescited by 262
- 2020IntelliCode Compose: Code Generation Using Transformercited by 225
- 2018SequenceR: Sequence-to-Sequence Learning for End-to-End Program Repaircited by 144
- 2023InferFix: End-to-End Program Repair with LLMscited by 113
- 2018An Empirical Study on Learning Bug-Fixing Patches in the Wild via Neural Machine Translationcited by 106
- 2023Learning Performance-Improving Code Editscited by 103
- 2022Repair Is Nearly Generation: Multilingual Program Repair with LLMscited by 68
- 2018CODIT: Code Editing with Tree-Based Neural Modelscited by 40
- 2016Automated Clustering and Program Repair for Introductory Programming Assignmentscited by 21
- 2021Generating Bug-Fixes Using Pretrained Transformerscited by 18
- 2016Automated Correction for Syntax Errors in Programming Assignments using Recurrent Neural Networkscited by 13
- 2016sk_p: a neural program corrector for MOOCscited by 10
- 2018Learning to Generate Corrective Patches using Neural Machine Translationcited by 10
- 2021FAPR: Fast and Accurate Program Repair for Introductory Programming Coursescited by 4
Built on by
Citing papers per year: 2023: 2.
2 papers.