arXiv subjects

Sungmin Kang

Publications and source records attributed to Sungmin Kang.

30 records · Page 2Linked to original sources

arXiv2024

Identifying Inaccurate Descriptions in LLM-generated Code Comments via Test Execution

Software comments are critical for human understanding of software, and as such many comment generation techniques have been proposed. However, we find that a systematic evaluation of the factual accuracy of generated comments is rare; only subjective accuracy labels have been given. Evaluating comments generated by three Large Language Models (LLMs), we find that even for the best-performing LLM, roughly a fifth of its comments contained demonstrably inaccurate statements. While it seems code-comment consistency detection techniques should be able to detect inaccurate comments, we perform experiments demonstrating they have no statistically significant relationship with comment accuracy, underscoring the substantial difficulty of this problem. To tackle this, we propose the concept of document testing, in which a document is verified by using an LLM to generate tests based on the document, running those tests, and observing whether they pass or fail. Furthermore, we implement our concept to verify Java comments. Experiments demonstrate that our approach has a robust statistical relationship with comment accuracy, making headway into a problem where prior techniques failed. Qualitative evaluation also reveals the promise of our approach in gaining developer trust, while highlighting the limitations of our current implementation.

cs.SE↗

arXiv2023

Evaluating Diverse Large Language Models for Automatic and General Bug Reproduction

Bug reproduction is a critical developer activity that is also challenging to automate, as bug reports are often in natural language and thus can be difficult to transform to test cases consistently. As a result, existing techniques mostly focused on crash bugs, which are easier to automatically detect and verify. In this work, we overcome this limitation by using large language models (LLMs), which have been demonstrated to be adept at natural language processing and code generation. By prompting LLMs to generate bug-reproducing tests, and via a post-processing pipeline to automatically identify promising generated tests, our proposed technique LIBRO could successfully reproduce about one-third of all bugs in the widely used Defects4J benchmark. Furthermore, our extensive evaluation on 15 LLMs, including 11 open-source LLMs, suggests that open-source LLMs also demonstrate substantial potential, with the StarCoder LLM achieving 70% of the reproduction performance of the closed-source OpenAI LLM code-davinci-002 on the large Defects4J benchmark, and 90% of performance on a held-out bug dataset likely not part of any LLM's training data. In addition, our experiments on LLMs of different sizes show that bug reproduction using LIBRO improves as LLM size increases, providing information as to which LLMs can be used with the LIBRO pipeline.

cs.SE↗

arXiv2023

The GitHub Recent Bugs Dataset for Evaluating LLM-based Debugging Applications

Large Language Models (LLMs) have demonstrated strong natural language processing and code synthesis capabilities, which has led to their rapid adoption in software engineering applications. However, details about LLM training data are often not made public, which has caused concern as to whether existing bug benchmarks are included. In lieu of the training data for the popular GPT models, we examine the training data of the open-source LLM StarCoder, and find it likely that data from the widely used Defects4J benchmark was included, raising the possibility of its inclusion in GPT training data as well. This makes it difficult to tell how well LLM-based results on Defects4J would generalize, as for any results it would be unclear whether a technique's performance is due to LLM generalization or memorization. To remedy this issue and facilitate continued research on LLM-based SE, we present the GitHub Recent Bugs (GHRB) dataset, which includes 76 real-world Java bugs that were gathered after the OpenAI data cut-off point.

cs.SE↗

arXiv2023

Towards Autonomous Testing Agents via Conversational Large Language Models

Software testing is an important part of the development cycle, yet it requires specialized expertise and substantial developer effort to adequately test software. Recent discoveries of the capabilities of large language models (LLMs) suggest that they can be used as automated testing assistants, and thus provide helpful information and even drive the testing process. To highlight the potential of this technology, we present a taxonomy of LLM-based testing agents based on their level of autonomy, and describe how a greater level of autonomy can benefit developers in practice. An example use of LLMs as a testing assistant is provided to demonstrate how a conversational framework for testing can help developers. This also highlights how the often criticized hallucination of LLMs can be beneficial for testing. We identify other tangible benefits that LLM-driven testing agents can bestow, and also discuss potential limitations.

cs.SE↗

arXiv2023

Large Language Models are Few-shot Testers: Exploring LLM-based General Bug Reproduction

Many automated test generation techniques have been developed to aid developers with writing tests. To facilitate full automation, most existing techniques aim to either increase coverage, or generate exploratory inputs. However, existing test generation techniques largely fall short of achieving more semantic objectives, such as generating tests to reproduce a given bug report. Reproducing bugs is nonetheless important, as our empirical study shows that the number of tests added in open source repositories due to issues was about 28% of the corresponding project test suite size. Meanwhile, due to the difficulties of transforming the expected program semantics in bug reports into test oracles, existing failure reproduction techniques tend to deal exclusively with program crashes, a small subset of all bug reports. To automate test generation from general bug reports, we propose LIBRO, a framework that uses Large Language Models (LLMs), which have been shown to be capable of performing code-related tasks. Since LLMs themselves cannot execute the target buggy code, we focus on post-processing steps that help us discern when LLMs are effective, and rank the produced tests according to their validity. Our evaluation of LIBRO shows that, on the widely studied Defects4J benchmark, LIBRO can generate failure reproducing test cases for 33% of all studied cases (251 out of 750), while suggesting a bug reproducing test in first place for 149 bugs. To mitigate data contamination, we also evaluate LIBRO against 31 bug reports submitted after the collection of the LLM training data terminated: LIBRO produces bug reproducing tests for 32% of the studied bug reports. Overall, our results show LIBRO has the potential to significantly enhance developer efficiency by automatically generating tests from bug reports.

cs.SE↗

arXiv2023

Towards Objective-Tailored Genetic Improvement Through Large Language Models

While Genetic Improvement (GI) is a useful paradigm to improve functional and nonfunctional aspects of software, existing techniques tended to use the same set of mutation operators for differing objectives, due to the difficulty of writing custom mutation operators. In this work, we suggest that Large Language Models (LLMs) can be used to generate objective-tailored mutants, expanding the possibilities of software optimizations that GI can perform. We further argue that LLMs and the GI process can benefit from the strengths of one another, and present a simple example demonstrating that LLMs can both improve the effectiveness of the GI optimization process, while also benefiting from the evaluation steps of GI. As a result, we believe that the combination of LLMs and GI has the capability to significantly aid developers in optimizing their software.

cs.SE↗

arXiv2023

Explainable Automated Debugging via Large Language Model-driven Scientific Debugging

Automated debugging techniques have the potential to reduce developer effort in debugging, and have matured enough to be adopted by industry. However, one critical issue with existing techniques is that, while developers want rationales for the provided automatic debugging results, existing techniques are ill-suited to provide them, as their deduction process differs significantly from that of human developers. Inspired by the way developers interact with code when debugging, we propose Automated Scientific Debugging (AutoSD), a technique that given buggy code and a bug-revealing test, prompts large language models to automatically generate hypotheses, uses debuggers to actively interact with buggy code, and thus automatically reach conclusions prior to patch generation. By aligning the reasoning of automated debugging more closely with that of human developers, we aim to produce intelligible explanations of how a specific patch has been generated, with the hope that the explanation will lead to more efficient and accurate developer decisions. Our empirical analysis on three program repair benchmarks shows that AutoSD performs competitively with other program repair baselines, and that it can indicate when it is confident in its results. Furthermore, we perform a human study with 20 participants, including six professional developers, to evaluate the utility of explanations from AutoSD. Participants with access to explanations could judge patch correctness in roughly the same time as those without, but their accuracy improved for five out of six real-world bugs studied: 70% of participants answered that they wanted explanations when using repair tools, while 55% answered that they were satisfied with the Scientific Debugging presentation.

cs.SE↗

arXiv2022

A Bayesian Framework for Automated Debugging

Debugging takes up a significant portion of developer time. As a result, automated debugging techniques including Fault Localization (FL) and Automated Program Repair (APR) have garnered significant attention due to their potential to aid developers in debugging tasks. Despite intensive research on these subjects, we are unaware of a theoretic framework that highlights the principles behind automated debugging and allows abstract analysis of techniques. Such a framework would heighten our understanding of the endeavor and provide a way to formally analyze techniques and approaches. To this end, we first propose a Bayesian framework of understanding automated repair and find that in conjunction with a concrete statement of the objective of automated debugging, we can recover maximal fault localization formulae from prior work, as well as analyze existing APR techniques and their underlying assumptions. As a means of empirically demonstrating our framework, we further propose BAPP, a Bayesian Patch Prioritization technique that incorporates intermediate program values to analyze likely patch locations and repair actions, with its core equations being derived by our Bayesian framework. We find that incorporating program values allows BAPP to identify correct patches more precisely: when applied to the patches generated by kPAR, the rankings produced by BAPP reduce the number of required patch validation by 68% and consequently reduce the repair time by 34 minutes on average. Further, BAPP improves the precision of FL, increasing acc@5 on the studied bugs from 8 to 11. These results highlight the potential of value-cognizant automated debugging techniques, and further validates our theoretical framework. Finally, future directions that the framework suggests are provided.

cs.SE↗

arXiv2022

Arachne: Search Based Repair of Deep Neural Networks

The rapid and widespread adoption of Deep Neural Networks (DNNs) has called for ways to test their behaviour, and many testing approaches have successfully revealed misbehaviour of DNNs. However, it is relatively unclear what one can do to correct such behaviour after revelation, as retraining involves costly data collection and does not guarantee to fix the underlying issue. This paper introduces Arachne, a novel program repair technique for DNNs, which directly repairs DNNs using their input-output pairs as a specification. Arachne localises neural weights on which it can generate effective patches and uses Differential Evolution to optimise the localised weights and correct the misbehaviour. An empirical study using different benchmarks shows that Arachne can fix specific misclassifications of a DNN without reducing general accuracy significantly. On average, patches generated by Arachne generalise to 61.3% of unseen misbehaviour, whereas those by a state-of-the-art DNN repair technique generalise only to 10.2% and sometimes to none while taking tens of times more than Arachne. We also show that Arachne can address fairness issues by debiasing a gender classification model. Finally, we successfully apply Arachne to a text sentiment model to show that it generalises beyond Convolutional Neural Networks.

cs.LG↗

arXiv2022

GLAD: Neural Predicate Synthesis to Repair Omission Faults

Existing template and learning-based APR tools have successfully found patches for many benchmark faults. However, our analysis of existing results shows that omission faults pose a significant challenge to these techniques. For template based approaches, omission faults provide no location to apply templates to; for learning based approaches that formulate repair as Neural Machine Translation (NMT), omission faults similarly do not provide the faulty code to translate. To address these issues, we propose GLAD, a novel learning-based repair technique that specifically targets if-clause synthesis. GLAD does not require a faulty line as it is based on generative Language Models (LMs) instead of machine translation; consequently, it can repair omission faults. GLAD intelligently constrains the language model using a type-based grammar. Further, it efficiently reduces the validation cost by performing dynamic ranking of candidate patches using a debugger. Thanks to the shift from translation to synthesis, GLAD is highly orthogonal to existing techniques: GLAD can correctly fix 16 Defects4J v1.2 faults that previous NMT-based techniques could not, while maintaining a reasonable runtime cost, underscoring its utility as an APR tool and potential to complement existing tools in practice. An inspection of the bugs that GLAD fixes reveals that GLAD can quickly generate expressions that would be challenging for other techniques.

cs.SE↗

arXiv2020

SINVAD: Search-based Image Space Navigation for DNN Image Classifier Test Input Generation

The testing of Deep Neural Networks (DNNs) has become increasingly important as DNNs are widely adopted by safety critical systems. While many test adequacy criteria have been suggested, automated test input generation for many types of DNNs remains a challenge because the raw input space is too large to randomly sample or to navigate and search for plausible inputs. Consequently, current testing techniques for DNNs depend on small local perturbations to existing inputs, based on the metamorphic testing principle. We propose new ways to search not over the entire image space, but rather over a plausible input space that resembles the true training distribution. This space is constructed using Variational Autoencoders (VAEs), and navigated through their latent vector space. We show that this space helps efficiently produce test inputs that can reveal information about the robustness of DNNs when dealing with realistic tests, opening the field to meaningful exploration through the space of highly structured images.

cs.SE↗

arXiv2018

Interpreting Models by Allowing to Ask

Questions convey information about the questioner, namely what one does not know. In this paper, we propose a novel approach to allow a learning agent to ask what it considers as tricky to predict, in the course of producing a final output. By analyzing when and what it asks, we can make our model more transparent and interpretable. We first develop this idea to propose a general framework of deep neural networks that can ask questions, which we call asking networks. A specific architecture and training process for an asking network is proposed for the task of colorization, which is an exemplar one-to-many task and thus a task where asking questions is helpful in performing the task accurately. Our results show that the model learns to generate meaningful questions, asks difficult questions first, and utilizes the provided hint more efficiently than baseline models. We conclude that the proposed asking framework makes the learning agent reveal its weaknesses, which poses a promising new direction in developing interpretable and interactive models.

cs.AI↗