arXiv subjects

Zimin Chen

Publications and source records attributed to Zimin Chen.

14 recordsLinked to original sources

arXiv2025

Prometheus: Towards Long-Horizon Codebase Navigation for Repository-Level Problem Solving

Large Language Models (LLMs) have shown remarkable capabilities in automating software engineering tasks, spurring the emergence of coding agents that scaffold LLMs with external tools to resolve repository-level problems. However, existing agents still struggle to navigate large-scale codebases, as the Needle-in-a-Haystack problem persists even with million-token context windows, where relevant evidence is often overwhelmed by large volumes of irrelevant code and documentation. Prior codebase navigation approaches, including embedding-based retrieval, file-system exploration, and graph-based retrieval, address parts of this challenge but fail to capture the temporal continuity of agent reasoning, rendering agents stateless and causing repeated repository traversals that hinder scalable planning and reasoning. To address these limitations, we present Prometheus, a memory-centric coding agent framework for long-horizon codebase navigation. Prometheus represents the repository as a unified knowledge graph to encode semantic dependencies and employs a context engine augmented with working memory that retains and reuses previously explored contexts to ensure continuity across reasoning steps. Built upon this engine, Prometheus integrates memory-enhanced navigation into a multi-agent system for automated issue resolution, encompassing issue classification, bug reproduction, patch generation, and verification. Comprehensive experiments are conducted on two widely used issue resolution benchmarks, i.e., SWE-bench Verified and SWE-PolyBench Verified. Powered by GPT-5, Prometheus achieves state-of-the-art performance with 74.4% and 33.8% resolution rates on the two benchmarks, ranking Top-6 and Top-1 among open-source agent systems, respectively. Our data and code are available at https://github.com/EuniAI/Prometheus.

cs.SE↗

arXiv2024

PreciseBugCollector: Extensible, Executable and Precise Bug-fix Collection

Bug datasets are vital for enabling deep learning techniques to address software maintenance tasks related to bugs. However, existing bug datasets suffer from precise and scale limitations: they are either small-scale but precise with manual validation or large-scale but imprecise with simple commit message processing. In this paper, we introduce PreciseBugCollector, a precise, multi-language bug collection approach that overcomes these two limitations. PreciseBugCollector is based on two novel components: a) A bug tracker to map the codebase repositories with external bug repositories to trace bug type information, and b) A bug injector to generate project-specific bugs by injecting noise into the correct codebases and then executing them against their test suites to obtain test failure messages. We implement PreciseBugCollector against three sources: 1) A bug tracker that links to the national vulnerability data set (NVD) to collect general-wise vulnerabilities, 2) A bug tracker that links to OSS-Fuzz to collect general-wise bugs, and 3) A bug injector based on 16 injection rules to generate project-wise bugs. To date, PreciseBugCollector comprises 1057818 bugs extracted from 2968 open-source projects. Of these, 12602 bugs are sourced from bug repositories (NVD and OSS-Fuzz), while the remaining 1045216 project-specific bugs are generated by the bug injector. Considering the challenge objectives, we argue that a bug injection approach is highly valuable for the industrial setting, since project-specific bugs align with domain knowledge, share the same codebase, and adhere to the coding style employed in industrial projects.

cs.SE↗

arXiv2023

Supersonic: Learning to Generate Source Code Optimizations in C/C++

Software optimization refines programs for resource efficiency while preserving functionality. Traditionally, it is a process done by developers and compilers. This paper introduces a third option, automated optimization at the source code level. We present Supersonic, a neural approach targeting minor source code modifications for optimization. Using a seq2seq model, Supersonic is trained on C/C++ program pairs ($x_{t}$, $x_{t+1}$), where $x_{t+1}$ is an optimized version of $x_{t}$, and outputs a diff. Supersonic's performance is benchmarked against OpenAI's GPT-3.5-Turbo and GPT-4 on competitive programming tasks. The experiments show that Supersonic not only outperforms both models on the code optimization task but also minimizes the extent of the change with a model more than 600x smaller than GPT-3.5-Turbo and 3700x smaller than GPT-4.

cs.SE↗

arXiv2023

MuRS: Mutant Ranking and Suppression using Identifier Templates

Diff-based mutation testing is a mutation testing approach that only mutates lines affected by a code change under review. Google's mutation testing service integrates diff-based mutation testing into the code review process and continuously gathers developer feedback on mutants surfaced during code review. To enhance the developer experience, the mutation testing service implements a number of suppression rules, which target not-useful mutants-that is, mutants that have consistently received negative developer feedback. However, while effective, manually implementing suppression rules require significant engineering time. An automatic system to rank and suppress mutants would facilitate the maintenance of the mutation testing service. This paper proposes and evaluates MuRS, an automated approach that groups mutants by patterns in the source code under test and uses these patterns to rank and suppress future mutants based on historical developer feedback on mutants in the same group. To evaluate MuRS, we conducted an A/B testing study, comparing MuRS to the existing mutation testing service. Despite the strong baseline, which uses manually developed suppression rules, the results show a statistically significantly lower negative feedback ratio of 11.45% for MuRS versus 12.41% for the baseline. The results also show that MuRS is able to recover existing suppression rules implemented in the baseline. Finally, the results show that statement-deletion mutant groups received both the most positive and negative developer feedback, suggesting a need for additional context that can distinguish between useful and not-useful mutants in these groups. Overall, MuRS has the potential to substantially reduce the development and maintenance cost for an effective mutation testing service by automatically learning suppression rules.

cs.SE↗

arXiv2023

Learning the Relation between Code Features and Code Transforms with Structured Prediction

To effectively guide the exploration of the code transform space for automated code evolution techniques, we present in this paper the first approach for structurally predicting code transforms at the level of AST nodes using conditional random fields (CRFs). Our approach first learns offline a probabilistic model that captures how certain code transforms are applied to certain AST nodes, and then uses the learned model to predict transforms for arbitrary new, unseen code snippets. {Our approach involves a novel representation of both programs and code transforms. Specifically, we introduce the formal framework for defining the so-called AST-level code transforms and we demonstrate how the CRF model can be accordingly designed, learned, and used for prediction}. We instantiate our approach in the context of repair transform prediction for Java programs. Our instantiation contains a set of carefully designed code features, deals with the training data imbalance issue, and comprises transform constraints that are specific to code. We conduct a large-scale experimental evaluation based on a dataset of bug fixing commits from real-world Java projects. The results show that when the popular evaluation metric \emph{top-3} is used, our approach predicts the code transforms with an accuracy varying from 41\% to 53\% depending on the transforms. Our model outperforms two baselines based on history probability and neural machine translation (NMT), suggesting the importance of considering code structure in achieving good prediction accuracy. In addition, a proof-of-concept synthesizer is implemented to concretize some repair transforms to get the final patches. The evaluation of the synthesizer on the Defects4j benchmark confirms the usefulness of the predicted AST-level repair transforms in producing high-quality patches.

cs.SE↗

arXiv2022

Multimodal Representation for Neural Code Search

Semantic code search is about finding semantically relevant code snippets for a given natural language query. In the state-of-the-art approaches, the semantic similarity between code and query is quantified as the distance of their representation in the shared vector space. In this paper, to improve the vector space, we introduce tree-serialization methods on a simplified form of AST and build the multimodal representation for the code data. We conduct extensive experiments using a single corpus that is large-scale and multi-language: CodeSearchNet. Our results show that both our tree-serialized representations and multimodal learning model improve the performance of code search. Last, we define intuitive quantification metrics oriented to the completeness of semantic and syntactic information of the code data, to help understand the experimental findings.

cs.SE↗

arXiv2022

Neural Transfer Learning for Repairing Security Vulnerabilities in C Code

In this paper, we address the problem of automatic repair of software vulnerabilities with deep learning. The major problem with data-driven vulnerability repair is that the few existing datasets of known confirmed vulnerabilities consist of only a few thousand examples. However, training a deep learning model often requires hundreds of thousands of examples. In this work, we leverage the intuition that the bug fixing task and the vulnerability fixing task are related and that the knowledge learned from bug fixes can be transferred to fixing vulnerabilities. In the machine learning community, this technique is called transfer learning. In this paper, we propose an approach for repairing security vulnerabilities named VRepair which is based on transfer learning. VRepair is first trained on a large bug fix corpus and is then tuned on a vulnerability fix dataset, which is an order of magnitude smaller. In our experiments, we show that a model trained only on a bug fix corpus can already fix some vulnerabilities. Then, we demonstrate that transfer learning improves the ability to repair vulnerable C functions. We also show that the transfer learning model performs better than a model trained with a denoising task and fine-tuned on the vulnerability fixing task. To sum up, this paper shows that transfer learning works well for repairing security vulnerabilities in C compared to learning on a small dataset.

cs.SE↗

arXiv2021

A Software-Repair Robot based on Continual Learning

Software bugs are common and correcting them accounts for a significant part of costs in the software development and maintenance process. This calls for automatic techniques to deal with them. One promising direction towards this goal is gaining repair knowledge from historical bug fixing examples. Retrieving insights from software development history is particularly appealing with the constant progress of machine learning paradigms and skyrocketing `big' bug fixing data generated through Continuous Integration (CI). In this paper, we present R-Hero, a novel software repair bot that applies continual learning to acquire bug fixing strategies from continuous streams of source code changes, implemented for the single development platform Github/Travis CI. We describe R-Hero, our novel system for learning how to fix bugs based on continual training, and we uncover initial successes as well as novel research challenges for the community.

cs.SE↗

arXiv2019

Using Sequence-to-Sequence Learning for Repairing C Vulnerabilities

Software vulnerabilities affect all businesses and research is being done to avoid, detect or repair them. In this article, we contribute a new technique for automatic vulnerability fixing. We present a system that uses the rich software development history that can be found on GitHub to train an AI system that generates patches. We apply sequence-to-sequence learning on a big dataset of code changes and we evaluate the trained system on real world vulnerabilities from the CVE database. The result shows the feasibility of using sequence-to-sequence learning for fixing software vulnerabilities.

cs.SE↗

arXiv2019

Learning to Fix Build Errors with Graph2Diff Neural Networks

Professional software developers spend a significant amount of time fixing builds, but this has received little attention as a problem in automatic program repair. We present a new deep learning architecture, called Graph2Diff, for automatically localizing and fixing build errors. We represent source code, build configuration files, and compiler diagnostic messages as a graph, and then use a Graph Neural Network model to predict a diff. A diff specifies how to modify the code's abstract syntax tree, represented in the neural network as a sequence of tokens and of pointers to code locations. Our network is an instance of a more general abstraction that we call Graph2Tocopo, which is potentially useful in any development tool for predicting source code changes. We evaluate the model on a dataset of over 500k real build errors and their resolutions from professional developers. Compared to the approach of DeepDelta (Mesbah et al., 2019), our approach tackles the harder task of predicting a more precise diff but still achieves over double the accuracy.

cs.LG↗

arXiv2019

SequenceR: Sequence-to-Sequence Learning for End-to-End Program Repair

This paper presents a novel end-to-end approach to program repair based on sequence-to-sequence learning. We devise, implement, and evaluate a system, 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 it 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 950/4711 testing samples, and find correct patches for 14 bugs in Defects4J. It captures a wide range of repair operators without any domain-specific top-down design.

cs.SE↗

arXiv2019

The Remarkable Role of Similarity in Redundancy-based Program Repair

Recently, there have been original attempts to use the concept of "code similarity" in program repair, suggesting that similarity analysis has an important role in the repair process. However, there is no dedicated work to characterize and quantify the role of similarity in redundancy-based program repair, where the patch is composed from source code taken from somewhere else. This is where our paper makes a major contribution: we perform a deep and systematic analysis of the role of code similarity during the exploration of the repair search space. We define and set up a large-scale experiment based on four code similarity metrics that capture different similarities: character, token, semantic and structure similarity. Overall, we have computed 56 million similarity score over 15 million source code components. We show that with similarity analysis, at least 90% of search space can be ignored to find the correct patch. Code similarity is capable of ranking the correct repair ingredient first in 4 - 33 % of the considered cases.

cs.SE↗

arXiv2019

A Literature Study of Embeddings on Source Code

Natural language processing has improved tremendously after the success of word embedding techniques such as word2vec. Recently, the same idea has been applied on source code with encouraging results. In this survey, we aim to collect and discuss the usage of word embedding techniques on programs and source code. The articles in this survey have been collected by asking authors of related work and with an extensive search on Google Scholar. Each article is categorized into five categories: 1. embedding of tokens 2. embedding of functions or methods 3. embedding of sequences or sets of method calls 4. embedding of binary code 5. other embeddings. We also provide links to experimental data and show some remarkable visualization of code embeddings. In summary, word embedding has been successfully applied on different granularities of source code. With access to countless open-source repositories, we see a great potential of applying other data-driven natural language processing techniques on source code in the future.

cs.LG↗

arXiv2018

The CodRep Machine Learning on Source Code Competition

CodRep is a machine learning competition on source code data. It is carefully designed so that anybody can enter the competition, whether professional researchers, students or independent scholars, without specific knowledge in machine learning or program analysis. In particular, it aims at being a common playground on which the machine learning and the software engineering research communities can interact. The competition has started on April 14th 2018 and has ended on October 14th 2018. The CodRep data is hosted at https://github.com/KTH/CodRep-competition/.

cs.SE↗