Search

The use of Large Language Models (LLMs) in automated test generation is gaining popularity, with much of the research focusing on metrics like compilability rate, code coverage and bug detection. However, an equally important quality metric is the presence of test smells design flaws or anti patterns in test code that hinder maintainability and readability. In this study, we explore the diffusion of test smells in LLM generated unit test suites and compare them to those found in human written ones. We analyze a benchmark of 20,500 LLM-generated test suites produced by four models (GPT-3.5, GPT-4, Mistral 7B, and Mixtral 8x7B) across five prompt engineering techniques, alongside a dataset of 780,144 human written test suites from 34,637 projects. Leveraging TsDetect, a state of the art tool capable of detecting 21 different types of test smells, we identify and analyze the prevalence and co-occurrence of various test smells in both human written and LLM-generated test suites. Our findings reveal new insights into the strengths and limitations of LLMs in test generation. First, regarding prevalence, we observe that LLMs frequently generate tests with common test smells, such as Magic Number Test and Assertion Roulette. Second, in terms of co occurrence, certain smells, like Long Test and Useless Test, tend to co occur in LLM-generated suites, influenced by specific prompt techniques. Third, we find that project complexity and LLM specific factors, including model size and context length, significantly affect the prevalence of test smells. Finally, the patterns of test smells in LLM-generated tests often mirror those in human-written tests, suggesting potential data leakage from training datasets. These insights underscore the need to refine LLM-based test generation for cleaner code and suggest improvements in both LLM capabilities and software testing practices.

Bimodal software analysis initially appeared to be within reach with the advent of large language models. Unfortunately, the complex interplay of natural language text and code in software engineering, presents unique challenges that prevent pretrained models to generalize to a variety of tasks. We postulate that in-context learning for the code-text bimodality is a promising avenue. This paper thus introduces a comprehensive study of in-context code-text learning, focusing on leveraging pretrained CodeLLAMA models. We consider a diverse dataset encompassing 23 software engineering tasks, which we transform in an in-context learning format. To effectively extract informative features, we propose a configurable prompt template. Our proposed pipeline, InCTRL, then unifies prompt learning across various software engineering tasks. Extensive evaluation on the study datasets demonstrates the superiority of INCTRL-models in few-shot performance, surpassing state-of-the-art models including the support model, CodeLLAMA. Typically, we observe that applied to the CodeLLAMA model, INCTRL brings improvements in terms of precision (at least about 12\%) and recall (up to 93.88\%) on various tasks. For example, on the task of program repair, INCTRL improves the BLEU score of CodeLLAMA by 85 points, while for clone detection, INCTRL achieves an improvement of 69 percentage points. Moreover, INCTRL-models offer state-of-the-art performance when using retrieval-augmented generation on individual downstream tasks. Finally, we qualitatively analyze the benefits of INCTRL over CodeLLAMA and open-source all models for broader impact. We make our code and dataset publicly available at: \begin{center} {\url{https://anonymous.4open.science/r/inctrl-B65B}} \end{center}

This paper presents VisLingInstruct, a novel approach to advancing Multi-Modal Language Models (MMLMs) in zero-shot learning. Current MMLMs show impressive zero-shot abilities in multi-modal tasks, but their performance depends heavily on the quality of instructions. VisLingInstruct tackles this by autonomously evaluating and optimizing instructional texts through In-Context Learning, improving the synergy between visual perception and linguistic expression in MMLMs. Alongside this instructional advancement, we have also optimized the visual feature extraction modules in MMLMs, further augmenting their responsiveness to textual content. Our comprehensive experiments on MMLMs, based on FlanT5 and Vicuna, show that VisLingInstruct significantly improves zero-shot performance in visual multi-modal tasks. Notably, it achieves a 13.1% and 9% increase in accuracy over the prior state-of-the-art on the TextVQA and HatefulMemes datasets. Our main code is available at https://github.com/Zhudongsheng75/VisLingInstruct., Comment: Accepted to NAACL2024 main conference

Code review, which aims at ensuring the overall quality and reliability of software, is a cornerstone of software development. Unfortunately, while crucial, Code review is a labor-intensive process that the research community is looking to automate. Existing automated methods rely on single input-output generative models and thus generally struggle to emulate the collaborative nature of code review. This work introduces \tool{}, a novel multi-agent Large Language Model (LLM) system for code review automation. CodeAgent incorporates a supervisory agent, QA-Checker, to ensure that all the agents' contributions address the initial review question. We evaluated CodeAgent on critical code review tasks: (1) detect inconsistencies between code changes and commit messages, (2) identify vulnerability introductions, (3) validate code style adherence, and (4) suggest code revision. The results demonstrate CodeAgent's effectiveness, contributing to a new state-of-the-art in code review automation. Our data and code are publicly available (\url{https://github.com/Code4Agent/codeagent}).

Text-to-SQL, the task of translating natural language questions into SQL queries, is part of various business processes. Its automation, which is an emerging challenge, will empower software practitioners to seamlessly interact with relational databases using natural language, thereby bridging the gap between business needs and software capabilities. In this paper, we consider Large Language Models (LLMs), which have achieved state of the art for various NLP tasks. Specifically, we benchmark Text-to-SQL performance, the evaluation methodologies, as well as input optimization (e.g., prompting). In light of the empirical observations that we have made, we propose two novel metrics that were designed to adequately measure the similarity between SQL queries. Overall, we share with the community various findings, notably on how to select the right LLM on Text-to-SQL tasks. We further demonstrate that a tree-based edit distance constitutes a reliable metric for assessing the similarity between generated SQL queries and the oracle for benchmarking Text2SQL approaches. This metric is important as it relieves researchers from the need to perform computationally expensive experiments such as executing generated queries as done in prior works. Our work implements financial domain use cases and, therefore contributes to the advancement of Text2SQL systems and their practical adoption in this domain., Comment: 10 pages, ICSE-SEIP 2024

The target of Electronic Health Record (EHR) coding is to find the diagnostic codes according to the EHRs. In previous research, researchers have preferred to do multi-classification on the EHR coding task; most of them encode the EHR first and then process it to get the probability of each code based on the EHR representation. However, the question of complicating diseases is neglected among all these methods. In this paper, we propose a novel EHR coding framework, which is the first attempt at detecting complicating diseases, called Copy Recurrent Neural Network Structure Network (CRNNet). This method refers to the idea of adversarial learning; a Path Generator and a Path Discriminator are designed to more efficiently finish the task of EHR coding. We propose a copy module to detect complicating diseases; by the proposed copy module and the adversarial learning strategy, we identify complicating diseases efficiently. Extensive experiments show that our method achieves a 57.30\% ratio of complicating diseases in predictions, demonstrating the effectiveness of our proposed model. According to the ablation study, the proposed copy mechanism plays a crucial role in detecting complicating diseases., Comment: arXiv admin note: text overlap with arXiv:2305.13250

In the face of growing vulnerabilities found in open-source software, the need to identify {discreet} security patches has become paramount. The lack of consistency in how software providers handle maintenance often leads to the release of security patches without comprehensive advisories, leaving users vulnerable to unaddressed security risks. To address this pressing issue, we introduce a novel security patch detection system, LLMDA, which capitalizes on Large Language Models (LLMs) and code-text alignment methodologies for patch review, data enhancement, and feature combination. Within LLMDA, we initially utilize LLMs for examining patches and expanding data of PatchDB and SPI-DB, two security patch datasets from recent literature. We then use labeled instructions to direct our LLMDA, differentiating patches based on security relevance. Following this, we apply a PTFormer to merge patches with code, formulating hybrid attributes that encompass both the innate details and the interconnections between the patches and the code. This distinctive combination method allows our system to capture more insights from the combined context of patches and code, hence improving detection precision. Finally, we devise a probabilistic batch contrastive learning mechanism within batches to augment the capability of the our LLMDA in discerning security patches. The results reveal that LLMDA significantly surpasses the start of the art techniques in detecting security patches, underscoring its promise in fortifying software maintenance.

In recent years, patch representation learning has emerged as a necessary research direction for exploiting the capabilities of machine learning in software generation. These representations have driven significant performance enhancements across a variety of tasks involving code changes. While the progress is undeniable, a common limitation among existing models is their specialization: they predominantly excel in either predictive tasks, such as security patch classification, or in generative tasks such as patch description generation. This dichotomy is further exacerbated by a prevalent dependency on potentially noisy data sources. Specifically, many models utilize patches integrated with Abstract Syntax Trees (AST) that, unfortunately, may contain parsing inaccuracies, thus acting as a suboptimal source of supervision. In response to these challenges, we introduce PATCH-CLIP, a novel pre-training framework for patches and natural language text. PATCH-CLIP deploys a triple-loss training strategy for 1) patch-description contrastive learning, which enables to separate patches and descriptions in the embedding space, 2) patch-description matching, which ensures that each patch is associated to its description in the embedding space, and 3) patch-description generation, which ensures that the patch embedding is effective for generation. These losses are implemented for joint learning to achieve good performance in both predictive and generative tasks involving patches. Empirical evaluations focusing on patch description generation, demonstrate that PATCH-CLIP sets new state of the art performance, consistently outperforming the state-of-the-art in metrics like BLEU, ROUGE-L, METEOR, and Recall., Comment: The paper is incomplete, causing much confusion for the community

Patch representation is crucial in automating various software engineering tasks, like determining patch accuracy or summarizing code changes. While recent research has employed deep learning for patch representation, focusing on token sequences or Abstract Syntax Trees (ASTs), they often miss the change's semantic intent and the context of modified lines. To bridge this gap, we introduce a novel method, Patcherizer. It delves into the intentions of context and structure, merging the surrounding code context with two innovative representations. These capture the intention in code changes and the intention in AST structural modifications pre and post-patch. This holistic representation aptly captures a patch's underlying intentions. Patcherizer employs graph convolutional neural networks for structural intention graph representation and transformers for intention sequence representation. We evaluated Patcherizer's embeddings' versatility in three areas: (1) Patch description generation, (2) Patch accuracy prediction, and (3) Patch intention identification. Our experiments demonstrate the representation's efficacy across all tasks, outperforming state-of-the-art methods. For example, in patch description generation, Patcherizer excels, showing an average boost of 19.39% in BLEU, 8.71% in ROUGE-L, and 34.03% in METEOR scores.

Within the realm of advanced code retrieval, existing methods have primarily relied on intricate matching and attention-based mechanisms. However, these methods often lead to computational and memory inefficiencies, posing a significant challenge to their real-world applicability. To tackle this challenge, we propose a novel approach, the Hyperbolic Code QA Matching (HyCoQA). This approach leverages the unique properties of Hyperbolic space to express connections between code fragments and their corresponding queries, thereby obviating the necessity for intricate interaction layers. The process commences with a reimagining of the code retrieval challenge, framed within a question-answering (QA) matching framework, constructing a dataset with triple matches characterized as \texttt{<}negative code, description, positive code\texttt{>}. These matches are subsequently processed via a static BERT embedding layer, yielding initial embeddings. Thereafter, a hyperbolic embedder transforms these representations into hyperbolic space, calculating distances between the codes and descriptions. The process concludes by implementing a scoring layer on these distances and leveraging hinge loss for model training. Especially, the design of HyCoQA inherently facilitates self-organization, allowing for the automatic detection of embedded hierarchical patterns during the learning phase. Experimentally, HyCoQA showcases remarkable effectiveness in our evaluations: an average performance improvement of 3.5\% to 4\% compared to state-of-the-art code retrieval techniques.

The growth of open-source software has increased the risk of hidden vulnerabilities that can affect downstream software applications. This concern is further exacerbated by software vendors' practice of silently releasing security patches without explicit warnings or common vulnerability and exposure (CVE) notifications. This lack of transparency leaves users unaware of potential security threats, giving attackers an opportunity to take advantage of these vulnerabilities. In the complex landscape of software patches, grasping the nuanced semantics of a patch is vital for ensuring secure software maintenance. To address this challenge, we introduce a multilevel Semantic Embedder for security patch detection, termed MultiSEM. This model harnesses word-centric vectors at a fine-grained level, emphasizing the significance of individual words, while the coarse-grained layer adopts entire code lines for vector representation, capturing the essence and interrelation of added or removed lines. We further enrich this representation by assimilating patch descriptions to obtain a holistic semantic portrait. This combination of multi-layered embeddings offers a robust representation, balancing word complexity, understanding code-line insights, and patch descriptions. Evaluating MultiSEM for detecting patch security, our results demonstrate its superiority, outperforming state-of-the-art models with promising margins: a 22.46\% improvement on PatchDB and a 9.21\% on SPI-DB in terms of the F1 metric.

Commit message generation (CMG) is a challenging task in automated software engineering that aims to generate natural language descriptions of code changes for commits. Previous methods all start from the modified code snippets, outputting commit messages through template-based, retrieval-based, or learning-based models. While these methods can summarize what is modified from the perspective of code, they struggle to provide reasons for the commit. The correlation between commits and issues that could be a critical factor for generating rational commit messages is still unexplored. In this work, we delve into the correlation between commits and issues from the perspective of dataset and methodology. We construct the first dataset anchored on combining correlated commits and issues. The dataset consists of an unlabeled commit-issue parallel part and a labeled part in which each example is provided with human-annotated rational information in the issue. Furthermore, we propose \tool (\underline{Ex}traction, \underline{Gro}unding, \underline{Fi}ne-tuning), a novel paradigm that can introduce the correlation between commits and issues into the training phase of models. To evaluate whether it is effective, we perform comprehensive experiments with various state-of-the-art CMG models. The results show that compared with the original models, the performance of \tool-enhanced models is significantly improved., Comment: ASE2023 accepted paper

Electronic Health Record (EHR) coding involves automatically classifying EHRs into diagnostic codes. While most previous research treats this as a multi-label classification task, generating probabilities for each code and selecting those above a certain threshold as labels, these approaches often overlook the challenge of identifying complex diseases. In this study, our focus is on detecting complication diseases within EHRs. We propose a novel coarse-to-fine ICD path generation framework called the Copy Recurrent Neural Network Structure Network (CRNNet), which employs a Path Generator (PG) and a Path Discriminator (PD) for EHR coding. By using RNNs to generate sequential outputs and incorporating a copy module, we efficiently identify complication diseases. Our method achieves a 57.30\% ratio of complex diseases in predictions, outperforming state-of-the-art and previous approaches. Additionally, through an ablation study, we demonstrate that the copy mechanism plays a crucial role in detecting complex diseases., Comment: Need modification

Recently, the ChatGPT LLM has received great attention: it can be used as a bot for discussing source code, prompting it to suggest changes, provide descriptions or even generate code. Typical demonstrations generally focus on existing benchmarks, which may have been used in model training (i.e., data leakage). To assess the feasibility of using an LLM as a useful assistant bot for programmers, we must assess its realistic capabilities on unseen problems as well as its capabilities on various tasks. In this paper, we present an empirical study of ChatGPT's potential as a fully automated programming assistant, focusing on the tasks of code generation, program repair, and code summariziation. The study investigates ChatGPT's performance on common programming problems and compares it with state-of-the-art approaches on two benchmarks. Among several findings, our study shows that ChatGPT is effective in dealing with common programming problems. However, our experiments also reveal limitations in terms of its attention span: detailed descriptions will constrain the focus of ChatGPT and prevent it from leveraging its vast knowledge to solve the actual problem. Surprisingly, we have identified the ability of ChatGPT to reason the original intention of the code. We expect future work to build on this insight for dealing with the open question of the oracle problem. Our findings contribute interesting insights to the development of LLMs for programming assistance, notably by demonstrating the importance of prompt engineering, and providing a better understanding of ChatGPT's practical applications for software engineering.

Although the great success of open-domain dialogue generation, unseen entities can have a large impact on the dialogue generation task. It leads to performance degradation of the model in the dialog generation. Previous researches used retrieved knowledge of seen entities as the auxiliary data to enhance the representation of the model. Nevertheless, logical explanation of unseen entities remains unexplored, such as possible co-occurrence or semantically similar words of them and their entity category. In this work, we propose an approach to address the challenge above. We construct a graph by extracting entity nodes in them, enhancing the representation of the context of the unseen entity with the entity's 1-hop surrounding nodes. Furthermore, We added the named entity tag prediction task to apply the problem that the unseen entity does not exist in the graph. We conduct our experiments on an open dataset Wizard of Wikipedia and the empirical results indicate that our approach outperforms the state-of-the-art approaches on Wizard of Wikipedia.

Software development teams generally welcome any effort to expose bugs in their code base. In this work, we build on the hypothesis that mobile apps from the same category (e.g., two web browser apps) may be affected by similar bugs in their evolution process. It is therefore possible to transfer the experience of one historical app to quickly find bugs in its new counterparts. This has been referred to as collaborative bug finding in the literature. Our novelty is that we guide the bug finding process by considering that existing bugs have been hinted within app reviews. Concretely, we design the BugRMSys approach to recommend bug reports for a target app by matching historical bug reports from apps in the same category with user app reviews of the target app. We experimentally show that this approach enables us to quickly expose and report dozens of bugs for targeted apps such as Brave (web browser app). BugRMSys's implementation relies on DistilBERT to produce natural language text embeddings. Our pipeline considers similarities between bug reports and app reviews to identify relevant bugs. We then focus on the app review as well as potential reproduction steps in the historical bug report (from a same-category app) to reproduce the bugs. Overall, after applying BugRMSys to six popular apps, we were able to identify, reproduce and report 20 new bugs: among these, 9 reports have been already triaged, 6 were confirmed, and 4 have been fixed by official development teams, respectively.

Recently, language representation techniques have achieved great performances in text classification. However, most existing representation models are specifically designed for English materials, which may fail in Chinese because of the huge difference between these two languages. Actually, few existing methods for Chinese text classification process texts at a single level. However, as a special kind of hieroglyphics, radicals of Chinese characters are good semantic carriers. In addition, Pinyin codes carry the semantic of tones, and Wubi reflects the stroke structure information, \textit{etc}. Unfortunately, previous researches neglected to find an effective way to distill the useful parts of these four factors and to fuse them. In our works, we propose a novel model called Moto: Enhancing Embedding with \textbf{M}ultiple J\textbf{o}int Fac\textbf{to}rs. Specifically, we design an attention mechanism to distill the useful parts by fusing the four-level information above more effectively. We conduct extensive experiments on four popular tasks. The empirical results show that our Moto achieves SOTA 0.8316 ($F_1$-score, 2.11\% improvement) on Chinese news titles, 96.38 (1.24\% improvement) on Fudan Corpus and 0.9633 (3.26\% improvement) on THUCNews.

Recently, neural language representation models pre-trained on large corpus can capture rich co-occurrence information and be fine-tuned in downstream tasks to improve the performance. As a result, they have achieved state-of-the-art results in a large range of language tasks. However, there exists other valuable semantic information such as similar, opposite, or other possible meanings in external knowledge graphs (KGs). We argue that entities in KGs could be used to enhance the correct semantic meaning of language sentences. In this paper, we propose a new method CKG: Dynamic Representation Based on \textbf{C}ontext and \textbf{K}nowledge \textbf{G}raph. On the one side, CKG can extract rich semantic information of large corpus. On the other side, it can make full use of inside information such as co-occurrence in large corpus and outside information such as similar entities in KGs. We conduct extensive experiments on a wide range of tasks, including QQP, MRPC, SST-5, SQuAD, CoNLL 2003, and SNLI. The experiment results show that CKG achieves SOTA 89.2 on SQuAD compared with SAN (84.4), ELMo (85.8), and BERT$_{Base}$ (88.5).

In this work, we propose a novel perspective to the problem of patch correctness assessment: a correct patch implements changes that "answer" to a problem posed by buggy behaviour. Concretely, we turn the patch correctness assessment into a Question Answering problem. To tackle this problem, our intuition is that natural language processing can provide the necessary representations and models for assessing the semantic correlation between a bug (question) and a patch (answer). Specifically, we consider as inputs the bug reports as well as the natural language description of the generated patches. Our approach, Quatrain, first considers state of the art commit message generation models to produce the relevant inputs associated to each generated patch. Then we leverage a neural network architecture to learn the semantic correlation between bug reports and commit messages. Experiments on a large dataset of 9135 patches generated for three bug datasets (Defects4j, Bugs.jar and Bears) show that Quatrain can achieve an AUC of 0.886 on predicting patch correctness, and recalling 93% correct patches while filtering out 62% incorrect patches. Our experimental results further demonstrate the influence of inputs quality on prediction performance. We further perform experiments to highlight that the model indeed learns the relationship between bug reports and code change descriptions for the prediction. Finally, we compare against prior work and discuss the benefits of our approach.

Representation learning of source code is essential for applying machine learning to software engineering tasks. Learning code representation from a multilingual source code dataset has been shown to be more effective than learning from single-language datasets separately, since more training data from multilingual dataset improves the model's ability to extract language-agnostic information from source code. However, existing multilingual training overlooks the language-specific information which is crucial for modeling source code across different programming languages, while only focusing on learning a unified model with shared parameters among different languages for language-agnostic information modeling. To address this problem, we propose MetaTPTrans, a meta learning approach for multilingual code representation learning. MetaTPTrans generates different parameters for the feature extractor according to the specific programming language type of the input code snippet, enabling the model to learn both language-agnostic and language-specific information with dynamic parameters in the feature extractor. We conduct experiments on the code summarization and code completion tasks to verify the effectiveness of our approach. The results demonstrate the superiority of our approach with significant improvements on state-of-the-art baselines., Comment: Accepted by AAAI 2023