代码大模型 / AI 编程

GANCoder: An Automatic Natural Language-to-Programming Language Translation Approach based on GAN

Yabing Zhu, Yanfeng Zhang, Huili Yang, Fangjing Wang

专题命中 代码生成 ：code generation(abstract)；分类 cs.CL、cs.LG

Comments 10pages, 4 figures

Journal ref NLPCC 2019: Natural Language Processing and Chinese Computing

LoopInvGen: A Loop Invariant Generator based on Precondition Inference

Saswat Padhi, Rahul Sharma, Todd Millstein

专题命中 代码生成 ：program synthesis(abstract)；分类 cs.LG、cs.PL

Comments Tool Description ( for technical details, see our PLDI paper at https://doi.org/10.1145/2908080.2908099 ), SyGuS-COMP'19 Competition Contribution, 4 pages

Insertion-based Decoding with automatically Inferred Generation Order

Jiatao Gu, Qi Liu, Kyunghyun Cho

专题命中 代码生成 ：code generation(abstract)；分类 cs.CL、cs.LG

Comments Camera ready. Accepted by TACL

SIF: A Framework for Solidity Code Instrumentation and Analysis

Chao Peng, Sefa Akca, Ajitha Rajan

专题命中 代码生成 ：code generation(abstract)；分类 cs.SE、cs.PL

Pinpointing Performance Inefficiencies in Java

Pengfei Su, Qingsen Wang, Milind Chabbi, Xu Liu

专题命中 代码生成 ：code generation(abstract)；分类 cs.SE、cs.PL

Comments This is a full-version of our ESEC/FSE'2019 paper

Learning to Infer Program Sketches

Maxwell Nye, Luke Hewitt, Joshua Tenenbaum, Armando Solar-Lezama

专题命中 代码生成 ：program synthesis(abstract)；分类 cs.AI、cs.LG

Comments Accepted to ICML 2019

Unsupervised Neural Generative Semantic Hashing

Casper Hansen, Christian Hansen, Jakob Grue Simonsen, Stephen Alstrup, Christina Lioma

专题命中 代码生成 ：code generation(abstract)；分类 cs.CL、cs.LG

Comments SIGIR 2019

Programming with Applicative-like expressions

Jan Malakhovski, Sergei Soloviev

专题命中 代码生成 ：code generation(abstract)；分类 cs.SE、cs.PL

Comments 18 pages

"Why did you do that?": Explaining black box models with Inductive Synthesis

Görkem Paçacı, David Johnson, Steve McKeever, Andreas Hamfelt

专题命中 代码生成 ：program synthesis(abstract)；分类 cs.AI、cs.LG

Comments 12 pages, 1 figure, accepted for publication at the Solving Problems with Uncertainties workshop as part of ICCS 2019, Faro, Portugal, June 12-14

Toward Imitating Visual Attention of Experts in Software Development Tasks

Yoshiharu Ikutani, Nishanth Koganti, Hideaki Hata, Takatomi Kubo, Kenichi Matsumoto

专题命中 代码生成 ：code generation(abstract)；分类 cs.SE、cs.AI

Comments 4 pages, EMIP 2019

Algorithms for Runtime Generation of Homogeneous Classes of Objects

Dmytro O. Terletskyi

专题命中 代码生成 ：code generation(abstract)；分类 cs.SE、cs.AI

Comments International Conference on Cyber Security and Computer Science (ICONCS'18), October 18-20, 2018, Safranbolu, Turkey

Inference Over Programs That Make Predictions

Yura Perov

专题命中 代码生成 ：program synthesis(abstract)；分类 cs.LG、cs.PL

Comments The International Conference on Probabilistic Programming, 2018

StructVAE: Tree-structured Latent Variable Models for Semi-supervised Semantic Parsing

Pengcheng Yin, Chunting Zhou, Junxian He, Graham Neubig

专题命中 代码生成 ：code generation(abstract)；分类 cs.CL、cs.LG

Comments ACL 2018

Comparative Study of Eight Formal Specifications of the Message Authenticator Algorithm

Hubert Garavel, Lina Marsso

专题命中 代码生成 ：code generation(abstract)；分类 cs.SE、cs.PL

Comments In Proceedings MARS/VPT 2018, arXiv:1803.08668

Journal ref EPTCS 268, 2018, pp. 41-87

Differentiable Functional Program Interpreters

John K. Feser, Marc Brockschmidt, Alexander L. Gaunt, Daniel Tarlow

专题命中 代码生成 ：program synthesis(abstract)；分类 cs.LG、cs.PL

Summary - TerpreT: A Probabilistic Programming Language for Program Induction

Alexander L. Gaunt, Marc Brockschmidt, Rishabh Singh, Nate Kushman, Pushmeet Kohli, Jonathan Taylor, Daniel Tarlow

专题命中 代码生成 ：program synthesis(abstract)；分类 cs.AI、cs.LG

Comments 7 pages, 2 figures, 4 tables in 1st Workshop on Neural Abstract Machines & Program Induction (NAMPI), @NIPS 2016

TerpreT: A Probabilistic Programming Language for Program Induction

Alexander L. Gaunt, Marc Brockschmidt, Rishabh Singh, Nate Kushman, Pushmeet Kohli, Jonathan Taylor, Daniel Tarlow

专题命中 代码生成 ：program synthesis(abstract)；分类 cs.AI、cs.LG

Comments 50 pages, 20 figures, 4 tables

Proceedings of the First Workshop on Program Transformation for Programmability in Heterogeneous Architectures

Salvador Tamarit, Julio Mariño, Guillermo Vigueras, Manuel Carro

专题命中 代码生成 ：code generation(abstract)；分类 cs.SE、cs.PL

A visual programming language for drawing and executing flowcharts

Drazen Lucanin, Ivan Fabek

专题命中 代码生成 ：code generation(abstract)；分类 cs.SE、cs.PL

Comments 6 pages, conference paper

Journal ref Proceedings of the 34th International Convention MIPRO, 23-27 May 2011, p. 1679-1684

JaCoText: A Pretrained Model for Java Code-Text Generation

Jessica López Espejel, Mahaman Sanoussi Yahaya Alassan, Walid Dahhane, El Hassane Ettifouri

专题命中 代码生成 ：code generation(abstract,comments)；分类 cs.CL

Comments International Conference on Code Generation and Implementation Volume: 17

Journal ref Espejel, J. L., Alassan, M. S. Y., Dahhane, W., & Ettifouri, E. H. (2023). JaCoText: A Pretrained Model for Java Code-Text Generation. International Journal of Computer and Systems Engineering, 17(2), 100-105

OpenSBLI: A framework for the automated derivation and parallel execution of finite difference solvers on a range of computer architectures

Christian T. Jacobs, Satya P. Jammy, Neil D. Sandham

专题命中 代码生成 ：code generation(abstract)；分类 cs.SE；repository(comments)

Comments Author accepted version, with a small amendment: the link in the "Code Availability" section has been updated, and now refers to the OpenSBLI source code repository on GitHub. Accepted for publication in the Journal of Computational Science on 8 November 2016

Journal ref Journal of Computational Science 18 (2017) 12-23

Grading the Graders: Verification Autonomy Levels (L0-L5) for LLM Reasoning

评估评估者：面向大语言模型推理的验证自主等级（L0-L5）

Yajie Yin

专题命中 代码生成 ：code generation(abstract)；分类 cs.CL

AI总结 该研究提出验证自主等级（VAL）元标准，解决验证领域的等级概念混淆，明确不同验证方案的完备性边界，相关代码与评估材料已公开。

Comments Code and data: https://github.com/1549080929-debug/math_agent Keywords: LLM verification; verification autonomy; completeness; ground truth; trustworthy AI Writing and implementation assisted by an AI language model; all experiments, data, and research decisions are the author's own

A Comprehensive Review of Large Language Models for Nanophotonics: From Surrogate Modeling to Autonomous Design

面向纳米光子学的大语言模型综合综述：从代理建模到自主设计

Huanshu Zhang, Kegeng Tang, Lei Kang, Sawyer D. Campbell, Zihao Wang, Douglas H. Werner

专题命中 代码生成 ：code generation(abstract)；分类 cs.LG

AI总结 该综述探讨大语言模型（LLMs）如何通过语义接口、代码生成及工具编排改进纳米光子学工作流，梳理相关方法的两类模式及跨学科应用，展望具备物理感知的多模态基础模型，推动AI从被动工具向主动科研合作者转变。

Comments Accepted for publication in Advanced Photonics

r2py: AI-Assisted Conversion of R Statistical Packages to Python

r2py：用于AI辅助将R统计包转换为Python的框架

Yufei Cai, Jun Li

专题命中 代码生成 ：code generation(abstract)；分类 cs.PL

AI总结 本文提出r2py，这是Claude Code环境中首个七阶段AI辅助方法，可将R包系统转换为数值保真的原生Python库，在KernSmooth包上实现的r2py_kernsmooth通过518项测试，有效数字一致性达6-10位。

Comments v2: substantially extended. Adds a second case study (rpart) reached through R's .Call() interface, and a five-phase prologue reconstructing the portion of R's C API the package uses so its original C compiles without R. v1 covered KernSmooth only. Title shortened

GxP-Agent: Process-DAG Topology for Reliable Clinical Trial Programming with LLM Agents

GxP-Agent：基于流程有向无环图拓扑结构的可靠临床试验编程大模型智能体

Jaime Yan

机构 * Harrisburg University of Science and Technology（哈里斯堡科技大学） 纠错

专题命中 代码生成 ：code generation(abstract)；分类 cs.AI

AI总结 该研究提出基于流程DAG拓扑的多智能体系统GxP-Agent，在临床试验编程任务上显著优于现有方法，实现了高结构匹配率，验证了编码领域流程知识为图拓扑的有效性。

Comments Preprint. 9 pages main text, 3 figures, plus references and appendix

Agentic Porting, Construction and Initial Verification and Validation of Libraries within the Open Source Unified TRAnsient Multi-Phase Advanced Reactor simulation Kit (Outram Park) Part I: Thermal Hydraulics

开源统一瞬态多相先进反应堆模拟工具包（Outram Park）中库的智能体迁移、构建及初步验证与确认 第一部分：热工水力

Theodore Kay Chen Ong, Ethan Yew Hoe Wong, Sicong Xiao

专题命中 代码生成 ：code generation(abstract)；分类 cs.SE

AI总结 本研究将OpenFOAM库智能体迁移至Rust语言的Outram-Foam库，开发了两相HEM壅塞流求解器，验证显示其吻合度良好，智能体编码可提升开源库开发效率，人类专业知识仍为关键。

Comments 78 pages, 21 figures, 17 tables

Automatically Generating ML Compiler Backends from Tensor Accelerator ISA Descriptions

从张量加速器ISA描述自动生成ML编译器后端

Devansh Jain, Akash Pardeshi, Marco Frigo, Kaustubh Khulbe, Krut Patel, Saatvik Lochan, Jai Arora, Charith Mendis

专题命中 代码生成 ：code generation(abstract)；分类 cs.PL

AI总结 本文提出ACT工具，可基于ISA描述自动生成张量加速器的XLA编译器后端，其生成的后端性能优于商用编译器，且提升了AWS Trainium的代码生成覆盖率。

Comments Accepted at OOPSLA 2026

Comparing the Quality of Code Generated by Vibe Coding Tools

比较 vibe 编码工具生成代码的质量

Gustavo da Mota, Kiev Gama

专题命中 代码生成 ：code generation(abstract)；分类 cs.SE

AI总结 本研究对比 Lovable、v0、Replit 三种 vibe 编码工具生成代码的结构质量，发现三者存在不同质量特征，选择工具需考量结构层面的权衡。

Competing at Every Price Point with Agentic Evolution over a Menu of LLMs

基于LLMs菜单的智能体进化，在每个价格点展开竞争

Andrew Borthwick

专题命中 代码生成 ：code generation(abstract)；分类 cs.AI

AI总结 本研究提出进化元智能体RoboPhD，通过对含9个LLM端点的菜单进行智能体进化，在DS-1000和PaperFindingBench两个任务上，除一个位置外占据所有帕累托前沿槽位，实现各价格点的竞争优势。

Detecting Contaminated Code-Generation Prompt Batches via Influence Functions

利用影响函数检测受污染的代码生成提示批次

Francesco Quinzan, Noor Munir, Yishun Lu, Stephen Roberts

机构 * The University of Oxford（牛津大学） 纠错

专题命中 代码生成 ：code generation(abstract)；分类 cs.LG

AI总结 提出与威胁模型无关的CodeSIFT方法，利用影响函数检测受污染的代码生成提示批次，在3B至7B参数的三个代码LLM上，中高注入率下AUROC达0.98，性能优于静态分析基线。