检索结果-内蒙古大学图书馆

Structured Chain-of-Thought Prompting for Code Generation

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Li, Jia Li, Ge Li, Yongmin Jin, Zhi Peking University Beijing China

Large Language Models (LLMs) (e.g., ChatGPT) have shown impressive performance in code generation. LLMs take prompts as inputs, and Chain-of-Thought (CoT) prompting is the state-of-the-art prompting technique. CoT prompting asks LLMs first to generate CoTs (i.e., intermediate natural language reasoning steps) and then output the code. However, CoT prompting is designed for natural language generation and has low accuracy in code generation. In this paper, we propose Structured CoTs (SCoTs) and present a novel prompting technique for code generation, named SCoT prompting. Our motivation is source code contains rich structural information and any code can be composed of three program structures (i.e., sequence, branch, and loop structures) [3]. Intuitively, structured intermediate reasoning steps make for structured source code. Thus, we ask LLMs to use program structures to build CoTs, obtaining SCoTs. Then, LLMs generate the final code based on SCoTs. Compared to CoT prompting, SCoT prompting explicitly constraints LLMs to think about how to solve requirements from the view of source code and further the performance of LLMs in code generation. We apply SCoT prompting to two LLMs (i.e., ChatGPT and Codex) and evaluate it on three benchmarks (i.e., HumanEval, MBPP, and MBCPP). (1) SCoT prompting outperforms the state-of-the-art baseline - CoT prompting by up to 13.79% in Pass@1. (2) Human evaluation shows human developers prefer programs from SCoT prompting. (3) SCoT prompting is robust to examples and achieves substantial improvements. Copyright © 2023, The Authors. All rights reserved.

关键词： computer programming languages

Brain Activations during programming Tasks: TD-NIRS and EEG Study 8

学校读者我要写书评

暂无评论

Brain Activations during Programming Tasks: TD-NIRS and EEG ...

European Conferences on Biomedical Optics - Diffuse Optical Spectroscopy and Imaging VIII

作者： Amendola, Caterina Calcagno, Alessandra Re, Rebecca Pirovano, Ileana Contini, Davide Spinelli, Lorenzo Couceiro, Ricardo Duraes, Joao Barbosa, Raul de Carvalho, Paulo Madeira, Henrique Bianchi, Anna Maria Torricelli, Alessandro Politecn Milan Dipartimento Fis Piazza Leonardo da Vinci 32 Milan Italy Politecn Milan Dept Elect Informat & Bioengn DEIB Milan Italy CNR ISt Foton & Nanotecnol IFN Piazza Leonardo da Vinci 32 Milan Italy CNR Ist Tecnol Biomed Via Fratelli Cervi 93 I-20090 Milan Italy Univ Coimbra CISUC Coimbra Portugal

Neural and cerebral hemodynamic activities of 16 programmers were monitored during programming tasks by simultaneous EEG and Time-Domain fNIRS measurements aiming at identifying cognitive and emotional states during c... 详细信息

ISBN: (数字)9781510647077

ISBN: (纸本)9781510647077;9781510647060

关键词： Electroencephalography computer programming Brain activation Hemodynamics Brain Near infrared spectroscopy Sensors computer programming languages Statistical analysis Visualization

18 Million Links in Commit Messages: Purpose, Evolution, and Decay

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Xiao, Tao Baltes, Sebastian Hata, Hideaki Treude, Christoph Kula, Raula Gaikovina Ishio, Takashi Matsumoto, Kenichi Nara Institute of Science and Technology Japan University of Adelaide Australia Shinshu University Japan University of Melbourne Australia

Commit messages contain diverse and valuable types of knowledge in all aspects of software maintenance and evolution. Links are an example of such knowledge. Previous work on "9.6 million links in source code comments" showed that links are prone to decay, become outdated, and lack bidirectional traceability. We conducted a large-scale study of 18,201,165 links from commits in 23,110 GitHub repositories to investigate whether they suffer the same fate. Results show that referencing external resources is prevalent and that the most frequent domains other than *** are the external domains of Stack Overflow and Google Code. Similarly, links serve as source code context to commit messages, with inaccessible links being frequent. Although repeatedly referencing links is rare (4%), 14% of links that are prone to evolve become unavailable over time;e.g., tutorials or articles and software homepages become unavailable over time. Furthermore, we find that 70% of the distinct links suffer from decay;the domains that occur the most frequently are related to Subversion repositories. We summarize that links in commits share the same fate as links in code, opening up avenues for future work. © 2023, CC BY.

关键词： computer programming languages

Learning Probabilistic Termination Proofs 33rd

学校读者我要写书评

暂无评论

Learning Probabilistic Termination Proofs

33rd International Conference on computer-Aided Verification (CAV)

作者： Abate, Alessandro Giacobbe, Mirco Roy, Diptarko Univ Oxford Oxford England

ISBN: (纸本)9783030816889;9783030816872

We present the first machine learning approach to the termination analysis of probabilistic programs. Ranking supermartingales (RSMs) prove that probabilistic programs halt, in expectation, within a finite number of steps. While previously RSMs were directly synthesised from source code, our method learns them from sampled execution traces. We introduce the neural ranking supermartingale: we let a neural network fit an RSM over execution traces and then we verify it over the source code using satisfiability modulo theories (SMT);if the latter step produces a counterexample, we generate from it new sample traces and repeat learning in a counterexample-guided inductive synthesis loop, until the SMT solver confirms the validity of the RSM. The result is thus a sound witness of probabilistic termination. Our learning strategy is agnostic to the source code and its verification counterpart supports the widest range of probabilistic single-loop programs that any existing tool can handle to date. We demonstrate the efficacy of our method over a range of benchmarks that include linear and polynomial programs with discrete, continuous, state-dependent, multi-variate, hierarchical distributions, and distributions with undefined moments.

关键词： computer programming languages

MetaOCaml Theory and Implementation

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Kiselyov, Oleg Tohoku University Japan

Quasi-quotation (or, code templates) has long been used as a convenient tool for code generation, commonly implemented as a pre-processing/translation into code-generation combinators. The original MetaOCaml was also based on such translation, done post type checking. BER MetaOCaml employs a significantly different, efficient (especially in version N114) translation integrated with type-checking, in the least intrusive way. This paper presents the integrated efficient translation for the first time. Copyright © 2023, The Authors. All rights reserved.

关键词： computer programming languages

Is this Snippet Written by ChatGPT? An Empirical Study with a CodeBERT-Based Classifier

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Nguyen, Phuong T. Di Rocco, Juri Di Sipio, Claudio Rubei, Riccardo Di Ruscio, Davide Di Penta, Massimiliano Università degli studi dell’Aquila Italy

Since its launch in November 2022, ChatGPT has gained popularity among users, especially programmers who use it as a tool to solve development problems. However, while offering a practical solution to programming problems, ChatGPT should be mainly used as a supporting tool (e.g., in software education) rather than as a replacement for the human being. Thus, detecting automatically generated source code by ChatGPT is necessary, and tools for identifying AI-generated content may need to be adapted to work effectively with source code. This paper presents an empirical study to investigate the feasibility of automated identification of AI-generated code snippets, and the factors that influence this ability. To this end, we propose a novel approach called GPTSniffer, which builds on top of CodeBERT to detect source code written by AI. The results show that GPTSniffer can accurately classify whether code is human-written or AI-generated, and outperforms two baselines, GPTZero and OpenAI Text Classifier. Also, the study shows how similar training data or a classification context with paired snippets helps to boost classification performances. Copyright © 2023, The Authors. All rights reserved.

关键词： computer programming languages

A Backend Platform for Supporting the Reproducibility of Computational Experiments

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Costa, Lázaro Barbosa, Susana Cunha, Jácome Faculty of Engineering University of Porto Portugal INESC TEC Portugal

In recent years, the research community has raised serious questions about the reproducibility of scientific work. In particular, since many studies include some kind of computing work (for instance, executing an algorithm encoded using some programming language), reproducibility is also a technological challenge, not only in computer science, but in most research domains. Replicability and computational reproducibility are not easy to achieve, not only because researchers have diverse proficiency in computing technologies, but also because of the variety of computational environments that can be used. Indeed, it is challenging to recreate the same environment using the same frameworks, code, programming languages, dependencies, and so on. In this work, we propose an Integrated Development Environment allowing the share, configuration, packaging and execution of an experiment by setting the code and data used and defining the programming languages, code, dependencies, databases, or commands to execute to achieve consistent results for each experiment. After the initial creation and configuration, the experiment can be executed any number of times, always producing exactly the same results. Furthermore, it allows the execution of the experiment by using a different associated dataset, and it can be possible to verify the reproducibility and replicability of the results. Our platform aims to allow researchers in any field to create a reproducibility package for their science that can be re-executed on any other computer. To evaluate our platform, we used it to reproduce 25 experiments extracted from published papers. We have been able to successfully reproduce 20 (80%) of these experiments achieving the results reported in such works with minimum effort, thus showing that our approach is effective. © 2023, CC BY.

关键词： computer programming languages

What Challenges Do Developers Face About Checked-in Secrets in Software Artifacts?

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Basak, Setu Kumar Neil, Lorenzo Reaves, Bradley Williams, Laurie North Carolina State University United States

Throughout 2021, GitGuardian’s monitoring of public GitHub repositories revealed a two-fold increase in the number of secrets (database credentials, API keys, and other credentials) exposed compared to 2020, accumulating more than six million secrets. To our knowledge, the challenges developers face to avoid checked-in secrets are not yet characterized. The goal of our paper is to aid researchers and tool developers in understanding and prioritizing opportunities for future research and tool automation for mitigating checked-in secrets through an empirical investigation of challenges and solutions related to checked-in secrets. We extract 779 questions related to checked-in secrets on Stack Exchange and apply qualitative analysis to determine the challenges and the solutions posed by others for each of the challenges. We identify 27 challenges and 13 solutions. The four most common challenges, in ranked order, are: (i) store/version of secrets during deployment;(ii) store/version of secrets in source code;(iii) ignore/hide of secrets in source code;and (iv) sanitize VCS history. The three most common solutions, in ranked order, are: (i) move secrets out of source code/version control and use template config file;(ii) secret management in deployment;and (iii) use local environment variables. Our findings indicate that the same solution has been mentioned to mitigate multiple challenges. However, our findings also identify an increasing trend in questions lacking accepted solutions substantiating the need for future research and tool automation on managing secrets. Copyright © 2023, The Authors. All rights reserved.

关键词： computer programming languages

U Owns the Code That Changes and How Marginal Owners Resolve Issues Slower in Low-Quality Source Code

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Borg, Markus Tornhill, Adam Mones, Enys CodeScene AB and Lund University Malmö Sweden CodeScene AB Malmö Sweden

[Context] Accurate time estimation is a critical aspect of predictable software engineering. Previous work shows that low source code quality increases the uncertainty in issue resolution times. [Objective] Our goal is to evaluate how developers' project experience and file ownership are related to issue resolution times. [Method] We mine 40 proprietary software repositories and conduct an observational study. Using CodeScene, we measure source code quality and active development time connected to Jira issues. [Results] Most source code changes are made by either a marginal or dominant code owner. Also, most changes to low-quality source code are made by developers with low levels of ownership. In low-quality source code, marginal owners need 45% more time for small changes, and 93% more time for large changes. [Conclusions] Collective code ownership is a popular target, but industry practice results in many dominant and marginal owners. Marginal owners are particularly hampered when working with low-quality source code, which leads to productivity losses. In codebases plagued by technical debt, newly onboarded developers will require more time to complete tasks. Copyright © 2023, The Authors. All rights reserved.

关键词： computer programming languages