This paper provides a comprehensive review of the literature concerning the utilization of Natural Language Processing (NLP) techniques, with a particular focus on transformer-based large language models (LLMs) trained using Big Code, within the domain of AI-assisted programming tasks. LLMs, augmented with software naturalness, have played a crucial role in facilitating AI-assisted programming applications, including code generation, code completion, code translation, code refinement, code summarization, defect detection, and clone detection. Notable examples of such applications include the GitHub Copilot powered by OpenAI’s Codex and DeepMind AlphaCode. This paper presents an overview of the major LLMs and their applications in downstream tasks related to AI-assisted programming. Furthermore, it explores the challenges and opportunities associated with incorporating NLP techniques with software naturalness in these applications, with a discussion on extending AI-assisted programming capabilities to Apple’s Xcode for mobile software development. This paper also presents the challenges of and opportunities for incorporating NLP techniques with software naturalness, empowering developers with advanced coding assistance and streamlining the software development process.

This paper studies how AI-assisted programming and large language models (LLM) improve software developers' ability via AI tools (LLM agents) like Github Copilot and Amazon CodeWhisperer, while integrating human feedback to enhance reinforcement learning (RLHF) with crowd-sourced computation to enhance text-to-code generation. Additionally, we demonstrate that our Bayesian optimization framework supports AI alignment in code generation by distributing the feedback collection burden, highlighting the value of collecting human feedback of good quality. Our empirical evaluations demonstrate the efficacy of this approach, showcasing how LLM agents can be effectively trained for improved text-to-code generation. Our Bayesian optimization framework can be designed for general domain-specific languages, promoting the alignment of large language model capabilities with human feedback in AI-assisted programming for code generation.

A theoretical method based on the induction equivalence principle and Green's functions is presented here to calculate the scattered electromagnetic fields from exterior plane building facades in urban environment. The reflection coefficient of the whole inhomogeneous structure can be calculated for different incidence angles. The derivation of appropriate Green's functions is presented. The method takes into account the large-scale discontinuities on the facade. It is simple to implement and fast to obtain results because of the absence of poles in the Green's functions expressions. The accuracy of the method is discussed comparing to a commercial software and physical optics as another method based on equivalent currents. The article concludes with a statistical study on the influence of concrete permittivity variation on the reflection coefficient of a concrete–glass building facade.

Since data science is traditionally an advanced field taught at the college or university level, introducing its concepts to K-12 students can present unique learning challenges. As educational environments increasingly adopt data science curricula for K-12 students, the need for scalable, personalized teaching tools becomes critical. While the integration of large language models (LLMs) in educational environments offers significant potential for scalability and automation, it is important to note that the generated language output may not always be highly suitable for K-12 students. In this paper, we introduce the DSRAG, a novel educational automatic feedback generation framework that leverages Retrieval-Augmented Generation (RAG) and cloud-based LLMs to provide automated and personalized feedback for K-12 students engaged in data science education. DSRAG employs Langchain question-answering and RAG systems to manage educational content and generate feedback on the top of GPT-4. We also demonstrate the framework's capability to simplify complex concepts and align its responses to be pedagogically appropriate and understandable for K-12 students.