Recent developments in ai-generated code are merely the latest in a series of challenges to traditional computer science education. aicode generators, along with the plethora of available code on the Internet and sit...
详细信息
ISBN:
(纸本)9798400707896
Recent developments in ai-generated code are merely the latest in a series of challenges to traditional computer science education. aicode generators, along with the plethora of available code on the Internet and sites that facilitate contract cheating, are a striking contrast to the heroic notion of programmers toiling away to create artisanal code from whole cloth. We need not interpret this to mean that more, potentially automated, policing of student assignments is necessary: automated policing of student work is already fraught with complications and ethical concerns. We argue that instructors should instead reconsider assessment design in their pedagogy in light of recent developments, with a focus on how students build knowledge, practice skills, and develop processes. How can these new tools support students and the way they learn, and support the way that computer scientists will work in the years to come? This is an opportunity to revisit how computer science is taught, how it is assessed, how we think about and present academic integrity, and the role of the computer scientist in general.
Educators are increasingly concerned about the usage of Large Language Models (LLMs) such as ChatGPT in programming education, particularly regarding the potential exploitation of imperfections in Artificial Intellige...
详细信息
ISBN:
(纸本)9798400704987
Educators are increasingly concerned about the usage of Large Language Models (LLMs) such as ChatGPT in programming education, particularly regarding the potential exploitation of imperfections in Artificial Intelligence generated Content (aiGC) Detectors for academic misconduct. In this paper, we present an empirical study where the LLM is examined for its attempts to bypass detection by aiGC Detectors. This is achieved by generating code in response to a given question using different variants. We collected a dataset comprising 5,069 samples, with each sample consisting of a textual description of a coding problem and its corresponding human-written Python solution codes. These samples were obtained from various sources, including 80 from Quescol, 3,264 from Kaggle, and 1,725 from Leet-code. From the dataset, we created 13 sets of code problem variant prompts, which were used to instruct ChatGPT to generate the outputs. Subsequently, we assessed the performance of five aiGC detectors. Our results demonstrate that existing aiGC Detectors perform poorly in distinguishing between human-written code and ai-generated code .
The increasing use of large language models (LLMs) such as Openai's ChatGPT and Google's Bard in the software development industry raise questions about the security of generatedcode. Our research evaluates J...
详细信息
ISBN:
(纸本)9798350375367
The increasing use of large language models (LLMs) such as Openai's ChatGPT and Google's Bard in the software development industry raise questions about the security of generatedcode. Our research evaluates Java, C, and Python code samples that were generated by these LLMs. In our investigation, we assessed the consistency of code samples generated by each LLM, characterized the security of generatedcode, and asked both LLMs to evaluate and fix the weaknesses of their own generatedcode as well as the code of the other LLM. Using 133 unique prompts from Google code Jam competitions, we produced 3,854 code samples across three distinct programming languages. We found that the code produced by these LLMs is frequently insecure and prone to weaknesses and vulnerabilities. This concerns human developers who must exercise caution while employing these LLMs.
Software plagiarism is the reuse of software code without proper attribution and in violation of software licensing agreements or copyright laws. With the popularity of open-source software and the rapid emergence of ...
详细信息
ISBN:
(纸本)9783031649530;9783031649547
Software plagiarism is the reuse of software code without proper attribution and in violation of software licensing agreements or copyright laws. With the popularity of open-source software and the rapid emergence of ai Large Language Models such as ChatGPT and Google Bard, the concerns of plagiarized ai-generated code have been rising. code attribution has been used to aid in the detection of software plagiarism cases. In this paper, we investigate the authorship of ai-generated code. We analyze the feasibility of code attribution approaches to verify authorship of source codegenerated by ai-based tools and investigate scenarios when plagiarized aicode can be identified. We perform an attribution analysis of an ai-generated source code on a large sample of programs written by software developers and generated by ChatGPT and Google Bard tools. We believe our work offers valuable insights for both academia and the software development community while contributing to the research in the authorship style of the fast-growing ai conversational models, ChatGPT and Bard.
To ensure that Large Language Models (LLMs) effectively support user productivity, they need to be adjusted. Existing code Readability (CR) models can guide this alignment. However, there are concerns about their rele...
详细信息
ISBN:
(纸本)9798400705861
To ensure that Large Language Models (LLMs) effectively support user productivity, they need to be adjusted. Existing code Readability (CR) models can guide this alignment. However, there are concerns about their relevance in modern software engineering since they often miss the developers' notion of readability and rely on outdated code. This research assesses existing Java CR models for LLM adjustments, measuring the correlation between their and developers' evaluations of ai-generated Java code. Using the Repertory Grid Technique with 15 developers, we identified 12 key code aspects influencing CR that were consequently assessed by 390 programmers when labeling 120 ai-generated snippets. Our findings indicate that when ai generates concise and executable code, it's often considered readable by CR models and developers. However, a limited correlation between these evaluations underscores the importance of future research on learning objectives for adjusting LLMs and on the aspects influencing CR evaluations included in predictive models.
暂无评论