SequenceR: Sequence-to-Sequence Learning for End-to-End Program Repair
By Zimin Chen, Steve Kommrusch, Michele Tufano, Louis-Noël Pouchet, Denys Poshyvanyk, Monperrus Martin. 2018
554 citations recorded by the source corpus.
This paper presents a novel end-to-end approach to program repair based on sequence-to-sequence learning. We devise, implement, and evaluate a technique, called SequenceR, for fixing bugs based on sequence-to-sequence learning on source code. This approach uses the copy mechanism to overcome the unlimited vocabulary problem that occurs with big code. Our system is data-driven; we train it on 35,578 samples, carefully curated from commits to open-source repositories. We evaluate SequenceR on 4,711 independent real bug fixes, as well on the Defects4J benchmark used in program repair research. SequenceR is able to perfectly predict the fixed line for 950/4,711 testing samples, and find correct patches for 14 bugs in Defects4J benchmark. SequenceR captures a wide range of repair operators without any domain-specific top-down design.
Subfields: Information Systems, Software, Program Analysis & Verification
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Information Systems
- Vincent J. Hellendoorn
- Chun Xia
- Lingming Zhang
- Charles Sutton
- Uri Alon
- Miltiadis Allamanis
- Prem Devanbu
- Baishakhi Ray
In Software
- Chun Xia
- Lingming Zhang
- Charles Sutton
- Vincent J. Hellendoorn
- Prem Devanbu
- Neel Sundaresan
- Alexey Svyatkovskiy
- Percy Liang
In Program Analysis & Verification
- Chun Xia
- Lingming Zhang
- Rafael-Michael Karampatsis
- Tien Nhut Nguyen
- Shaohua Wang
- Yi Li
- Eran Yahav
- Uri Alon
Citations in the corpus
- Program Synthesis with Large Language Models (2021, methodology): “Many learning methods have been proposed for program repair [Allamanis et al., 2018b, Tarlow et al., 2019, Hellendoorn et al., 2019, Dinella et al., 2019, Yasunaga and Liang, 2020, Chen et al., 2019b, Pradel and Sen, 2018].”
- Break-It-Fix-It: Unsupervised Learning for Program Repair (2021, background): “Several works learn to repair code from labeled datasets of source code edits made by programmers, e.g., error resolution records (Just et al., 2014; Chen et al., 2019; Mesbah et al., 2019; Bader et al., 2019; Tarlow et al., 2020; Ding et al., 2020).”
- Graph-based, Self-Supervised Program Repair from Diagnostic Feedback (2020, background): “Another line of work learns from labeled datasets of how programmers edit code (e.g. error resolution records) (Just et al., 2014; Chen et al., 2019).”
- InCoder: A Generative Model for Code Infilling and Synthesis (2022, methodology): “This includes methods to infer variable types [50, 66], generate unit tests [25], repair programs [30, 70, 20, 71], and verify program correctness [55].”
- Unified Pre-training for Program Understanding and Generation (2021, background): “, 2020), Program Repair (Chen et al., 2019; Chakraborty et al., 2020; Lutellier et al., 2020), Code Translation (Chen et al.”
- Unsupervised Translation of Programming Languages (2020, background): “Another line of work studied the applications of neural networks to code suggestion [2, 11, 34], or error detection [13, 18, 47].”
- Unit Test Case Generation with Transformers (2020, background): “Deep learning techniques have shown the potential of learning from real-world examples, and have been employed in several software engineering tasks, such as code completion [17], automated patch generation [18], [19], comment generation [20], and many others [21].”
- DOBF: A Deobfuscation Pre-Training Objective for Programming Languages (2021, background): “[2020], Kim et al. [2020], Svyatkovskoy et al. [2020], bug detection and code repair Allamanis et al.”
- Big Code != Big Vocabulary: Open-Vocabulary Models for Source Code (2020, methodology): “also used for program repair [20], and method naming [5].”
- METHODS2TEST: A dataset of focal methods mapped to test cases (2022, methodology): “Examples of such datasets include Defects4J [14], Bugs2Fix [27], and CodRep [5] for automated bug-fixing, NASA Metrics Data Program (MDP) [23], PROMISE [2], and the unified dataset [9] for defect prediction, Landfill [19] for code smell detection, and many others.”
- Can OpenAI's Codex Fix Bugs?: An evaluation on QuixBugs (2022, methodology): “A variety of different APR approaches have been proposed in the recent years: i) using genetic-programming (e.g., GenProg [8] or ARJA [29]), ii) using repair patterns (such as PAR [13], ELIXIR [24] or TBar [17]), iii) code retrieval-based approaches (e.g., ssFix [26] or LSRepair [18]), iv) or using ”
- Automated Program Repair in the Era of Large Pre-trained Language Models (2022, unknown): “We choose 8 recent learning-based APR tools: AlphaRepair [26], Re-wardRepair [23], Recoder [21], DeepDebug [65], CURE [22], CoCoNuT [24], DLFix [66] and SequenceR [67].”
- Less training, more repairing please: revisiting automated program repair via zero-shot learning (2022, methodology): “SequenceR [14] is a sequence-to-sequence network built using a Long Short-Term Memory (LSTM) [28] encoder and decoder network for program repair.”
- Patching as Translation: the Data and the Metaphor (2020, methodology): “[6] subsequently introduced SequenceR, an end-to-end framework to repair one-line Java bugs.”
- Baldur: Whole-Proof Generation and Repair with Large Language Models (2023, background): “, heuristic-based [36, 44, 48, 64, 82, 89], constraintbased [1, 23, 37, 51, 85], and learning-based [12, 24, 72]), and patch validationmethodologies [81, 86, 94, 98, 99].”