Software Engineering and Development
Applications of LLMs in improving software development processes, including code generation, debugging, testing, and specification development
Software Engineering and Development
Research on Large Language Models in Software Engineering and Development
AI-Powered Bug Explanation
Leveraging Code Structures to Help Developers Understand Software Bugs
Domain Expertise Trumps Model Size
Why medium-sized specialized AI models outperform larger generalists for software tasks
Leveraging LLMs for Software Specifications
Evaluating GPT-4's ability to generate formal software specifications from natural language
AI vs. Humans: Developer Coding Support Preferences
How developers engage with ChatGPT compared to Stack Overflow
Prompt Engineering vs. Fine-Tuning for Code
Comparing LLM approaches for code-related tasks
3D Vision-Language Model for Robotics
Advancing Robotic Vision Beyond Closed-Set Limitations
Domain-Specific Code Generation with LLMs
Evaluating LLM effectiveness beyond general programming tasks
Smarter Automated Testing with AI
Using LLMs to extract test inputs from bug reports
Modular Thinking for Advanced Coding
Breaking down complex programming problems with modular AI approaches
AI-Powered Specification Generation
Automating Formal Program Specifications with LLMs
Do Code Models Truly Understand Code?
Testing LLMs' comprehension of programming concepts using counterfactual analysis
API Pack: Supercharging Code Generation
Teaching LLMs to master API calls across programming languages
AI-Powered Test Case Generation
Enhancing LLMs for Test-Driven Development Automation
ChatDBG: Revolutionizing Software Debugging
How LLMs Transform Bug-Hunting Into Natural Conversation
AgentFL: Scaling Bug Detection to Project Level
LLM-powered fault localization for complex codebases
Bridging Design and Code with AI
WebCode2M: A breakthrough dataset revolutionizing automated webpage development
LLM-Powered Mutation Testing
Revolutionizing Test Quality Assessment with AI
Real-World Code Translation to Rust
First major study evaluating LLM performance on practical language migration
Smart Cascading for Code Generation
A cost-efficient framework that balances accuracy and computational resources
Boosting Automated Verification with LLMs
How AI can streamline software proof engineering
Teaching AI to Ask Better Questions
Making LLMs Better Coders Through Improved Communication
Fixing Python Library Compatibility Issues Automatically
Introducing PCART: An automated tool for resolving API parameter conflicts
Benchmarking LLMs for Test Case Generation
A systematic evaluation framework for LLM-powered software testing
Benchmarking LLM Code Efficiency
A new standard for evaluating AI-generated code quality
Evolution-Aware Code Generation Evaluation
Measuring LLMs on real-world software development dynamics
API Time Warp: LLMs Stuck in the Past?
Revealing how LLMs struggle with deprecated APIs in code completion
AI vs. Traditional Compilers for Code Optimization
Evaluating LLMs as the next frontier in code efficiency
Automated Test Generation with AI Agents
Validating Real-World Bug Fixes Using LLM-Based Code Agents
Boosting Code Generation with External Knowledge
Evaluating the Impact of Retrieval-Augmented Generation (RAG) for Programming Tasks
Beyond Simple Coding: Advancing AI Code Generation
New benchmark reveals capabilities and limitations of LLMs in complex programming tasks
Smart Self-Editing Code Assistants
Efficient, Real-Time Code Prediction During Editing
Keeping AI Code Assistants Up-to-Date
A new benchmark for evaluating how LLMs adapt to API changes
Scaling Code Generation Without Expensive Teachers
A new approach for creating synthetic programming data at scale
Evaluating LLMs for Software Development
A self-evaluating framework for measuring AI performance in coding tasks
Supercharging LLMs for Software Development
How Patched MOA optimizes inference to achieve superior performance
Evolving Better Code Instructions
How Genetic Algorithms Create High-Quality Training Data for LLMs
Measuring the True Difficulty of Code Tasks
A smarter approach to evaluating LLMs on programming challenges
TestART: Evolving LLM-based Unit Testing
Enhancing test quality through automated generation and repair cycles
LLMs: Breaking the Language Barrier in Code Clone Detection
Leveraging Large Language Models to identify similar code across programming languages
COAST: Revolutionizing Code Debugging in LLMs
A multi-agent approach to enhance debugging capabilities in AI coding systems
Combating Dockerfile Flakiness
The first systematic approach to detect and repair unpredictable Docker build failures
LLMs as Software Engineering Evaluators
Using AI to replace human annotation in software engineering research
Anchoring Attention in LLM Code Generation
Combating attention dilution for more accurate code
Evaluating Code Generation LLMs
A comprehensive framework for assessing AI code generation capabilities
LLMs and Buggy Code: Testing Implications
How incorrect code affects test generation by language models
Conversational Agents in Software Engineering
Maximizing Value of Bots in Development Workflows
AI-Powered Verification for Rust
Automating Correctness Proofs with Large Language Models
Zero-Shot Code Embedding with LLMs
Eliminating the need for supervised training in code analysis
Navigating the Challenges of LLM Open-Source Projects
Identifying key issues, causes, and practical solutions for practitioners
Teaching Code LLMs to Learn from Mistakes
Reinforcement Learning for Self-Improving AI Code Generation
ToolGen: Revolutionizing AI Tool Interaction
Embedding tool knowledge directly into language models
Smart Switching: Code vs. Text Reasoning in LLMs
Optimizing when LLMs should code rather than reason through text
Enhancing AI-Powered Code Repair
Improving robustness of LLM code repair systems through metamorphic testing
Optimizing RAG Systems for Code Generation
Balancing Effectiveness and Efficiency in Retrieval for Coding Tasks
SwiftCoder: Making AI-Generated Code Better & Faster
Fine-tuning LLMs to prioritize both correctness and efficiency in code generation
RepoGraph: Beyond Function-Level Code Generation
Repository-wide context for enhanced AI software engineering
CAMP: Bridging Local & Cloud Development
A hybrid AI coding assistant for constrained environments
Monte Carlo Tree Search for Smarter Software Agents
Enhancing LLM-based coding assistants with strategic decision-making capabilities
Beyond Cool Demos: Production-Ready AI Systems
Bridging the gap between prototype and enterprise-ready foundation model software
FALCON: Optimizing AI Code Generation
Reinforced learning system that adapts to user feedback in real-time
VisualCoder: Guiding Large Language Models in Code Execution...
By Cuong Chi Le, Hoang-Chau Truong-Vinh...
Beyond Line-by-Line Code Translation
A Repository-Level Approach Using Large Language Models
Bridging the Language Gap in Scientific Computing
Using LLMs to modernize legacy code with human oversight
DemoCraft: Smarter Code Generation with LLMs
Enhancing AI coding through in-context learning and latent concept awareness
Understanding LLM Code Generation Failures
Beyond Syntax: Why LLMs Struggle with Context-Dependent Programming
OpenCoder: Democratizing Code AI Development
A transparent, reproducible approach to building top-tier code LLMs
Transforming Chaos into Clarity
Visualizing Software Logs with Time Curves for Enhanced Security Analysis
Self-Correcting Code Generation with LLMs
Enhancing AI code quality through dynamic backtracking
Evaluating LLMs in Software Engineering
A systematic framework for rigorous empirical research
Building LLMs for Production: A Practitioner's Guide
Key considerations for developing enterprise-ready LLM applications
PyGen: AI-Powered Python Package Creation
Automating software development through human-AI collaboration
Open-Source LLMs for Smarter Bug Detection
Flexible Fault Localization Without Proprietary AI Dependencies
Benchmark Contamination in LLMs
Investigating data leakage in bug detection benchmarks
Beyond Function-Level: Repository-Scale Code Translation
Advancing LLMs for real-world Rust translation challenges
Improving Code Retrieval with Quality Data
A Contrastive Approach for Better Software Engineering
LLMs in Qualitative Research: A Software Engineering Revolution
Transforming how we analyze human and social factors in software development
Advancing Verilog Generation with LLMs
A hierarchical dataset approach to improve hardware design coding
Predicting Application Behavior with AI
Using Lightweight Generative Models to Optimize Software Performance
Do Top LLMs Excel in Specialized Coding Domains?
Evaluating code generation across real-world application domains
Order Bias in LLMs for Code Analysis
How input arrangement impacts fault localization accuracy
Modernizing Legacy Code with AI
Using LLMs to Transform Fortran to C++ through Intelligent Dialogue
Open-Source LLMs for Better GitHub Issue Resolution
A more accessible approach to fixing software bugs with AI
Smart LLM Selection for Program Synthesis
Dynamically choosing the optimal prompt and model for coding tasks
Probabilistic Analysis for LLM-Enabled Software
A framework for reliability and verification in AI systems
Smarter API Testing with Small LLMs
Targeted fine-tuning outperforms larger general-purpose models
Revolutionizing REST API Testing with AI
Smarter, more comprehensive API testing through LLMs and reinforcement learning
AI-Developer Collaboration Patterns
A taxonomy of how engineers interact with AI tools in software development
Smart Bug Hunting
Using AI feedback to pinpoint software bugs faster
Energy-Efficient AI Code Generation
Optimizing LLMs for sustainable software development
AI-Powered Chaos Engineering
Automating system resilience testing with LLMs
SkillScope: Matching Engineers to GitHub Issues
AI-driven skill prediction for open source contributions
Beyond Single Functions: Translating Entire Code Repositories
A new framework for evaluating large language models in code migration projects
AI-Powered Bug Fixing: A Collaborative Approach
Enhancing LLMs with programmer intent and collaborative behavior simulation
Smart C++ Test Generation with CITYWALK
Enhancing LLMs for Compiled Language Test Creation
AI-Powered Dependency Resolution
Using Large Language Models to Fix Python Dependency Conflicts
Mind the Gap: LLMs vs. Human Code Understanding
Revealing limitations in AI structural code comprehension
LLM Critics: Smarter Code Evaluation Without Execution
Using AI to assess code changes with unprecedented detail
Human vs. AI Code: A Critical Comparison
Evaluating LLMs against human programmers on real coding tasks
The Practitioner's View on Code-Generating LLMs
Bridging the gap between academic research and developer needs
Making Python Code More Pythonic with AI
Using LLMs to Automatically Refactor Non-Idiomatic Code
Smarter Test Validation with AI
Using LLMs to Generate Accurate Test Oracles
AI-Powered Crash Analysis
Localizing Software Faults with Large Language Models
LLM Program Optimization via Retrieval Augmented Search
By Sagnik Anupam, Alexander Shypula...
Bridging the Gap for Low-Resource Programming Languages
Improving LLMs' code generation capabilities for niche programming languages
Automating Legacy Web App Modernization
Using Multi-Agent LLM Systems to Transform Outdated Applications
LLMs vs. Human Experts in Requirements Engineering
Evaluating AI's potential to transform software requirements gathering
Evaluating LLMs for Real-World Coding Tasks
Beyond single-file solutions: Testing LLMs on multi-file project problems
Precision Bug Hunting with AI
How OrcaLoca enhances software issue localization with LLM agents
Bridging Code and Documentation
A new benchmark dataset for AI-powered software maintenance
How Students Use LLMs in Software Engineering Education
Insights from real-world student-AI interactions in the classroom
LLMs as Verification Partners
Leveraging AI to ensure Java code correctness
Smarter Unit Testing with AI
Training LLMs to generate tests that find and fix code bugs
Toward Neurosymbolic Program Comprehension
By Alejandro Velasco, Aya Garryyeva...
AI Agents for Automated Bug Fixing
Leveraging LLMs to reproduce bugs from incomplete reports at Google
SE Arena: Revolutionizing LLM Evaluation
An interactive platform for assessing AI models in software engineering workflows
Can LLMs Fix Real-World Code Maintenance Issues?
Evaluating Copilot Chat and Llama 3.1 on actual GitHub maintainability problems
AI-Driven ML Library Development
Self-improving LLMs that tackle specialized programming tasks
Automated Architectural Component Generation
Using LLMs to Generate Serverless Functions
Taming Flaky Tests with LLMs
Leveraging AI to detect and classify non-deterministic tests
Stable & Cost-Efficient AI Patching Framework
Balancing Efficiency and Accuracy in Automated Software Repairs
Improving AI Code Review Quality
Enhancing neural models by tackling noisy training data
Beyond Correctness: Benchmarking LLM Code Efficiency
Measuring and improving the time efficiency of AI-generated code
Evaluating LLMs for Test Case Generation
A systematic framework for assessing AI-powered software testing
SLM Ensembles: Smarter, Smaller AI for Code Debugging
Harnessing small language models to efficiently locate software bugs
Self-Debugging AI Code Generation
Two-agent framework improves automated code quality and reliability
AI-Powered Code Review Enhancement
Using LLMs to generate curated, actionable review comments
Efficient Code Summarization for Sustainable AI
Reducing computational costs while maintaining effectiveness
Automating Vehicle API Testing with LLMs
How AI can streamline the complete automotive test process
Automating Microservice Architecture Documentation
Using LLMs to detect patterns in Infrastructure-as-Code
AI-Powered Microservices Migration
Using Contrastive Learning to Revolutionize Software Architecture Transformation
Reimagining Software Agents
Accessing software internals for more accurate and efficient AI assistance
CodeSCM: Causal Analysis for Multi-Modal Code Generation
By Mukur Gupta, Noopur Bhatt...
LLM Agents for Research Deployment
Benchmarking AI assistants for complex code repositories
LLMs as Code Quality Judges
Can AI replace human evaluators in software engineering?
Benchmarking Data Leakage in LLMs
First comprehensive study across 83 software engineering benchmarks
Smarter Code Reviews with AI
Combining LLMs with Static Analysis for More Effective Code Feedback
Supercharging Competitive Programming with AI
How Large Language Models Excel at Olympic-Level Coding Challenges
LLMs' Understanding of Compiler Code
First empirical study on how well LLMs comprehend Intermediate Representations
Bridging Human Language and Command Line
Enhancing LLM-based translation from natural language to Bash
CodeQUEST: AI-Powered Code Quality Enhancement
Using LLMs to automatically evaluate and improve code across multiple dimensions
Verifying AI-Generated Code
Formal Verification Techniques for LLM Output
Enhancing AI-Generated Code Evaluation
A New Approach to Bridge Code and Requirements
Evaluating AI Coding Assistants
Bringing Human-Centered Design to Automated LLM Evaluation
Proactive Safety Engineering for ML Systems
Using LLMs to support hazard identification and mitigation in ML-powered applications
AI-Powered Code Refinement
Transforming Code Quality Through Intent-Driven Automation
From Sketch to Code: AI-Powered GIS Dashboards
Converting PowerPoint UI sketches to functional web applications using LLMs
Self-Improving Code Generation
How LLMs Can Refine Their Own Code Through Adaptive Critique
Automating Microservice Development with AI
Using LLMs to generate complete RESTful API implementations
LLMs Transform Unit Testing
Enhancing Defect Detection and Developer Efficiency
Rethinking LLM Programming
A Theoretical Framework Based on Automata Theory
Open-Source AI for Python Optimization
Democratizing performance profiling with open-source LLMs
Evolving Code Generation with CoCoEvo
A Novel Framework for Simultaneous Evolution of Programs and Test Cases
LintLLM: Revolutionizing Verilog Code Quality
Leveraging Large Language Models for Smarter Hardware Code Linting
LLMs as Code Executors
Transforming code analysis with LLM-powered surrogate models
ToolCoder: Teaching LLMs to Use Tools Through Code
Transforming tool learning into a structured code generation problem
Code Generation Reimagined
Using Unit Tests to Guide Automated Code Synthesis
Enhancing AI Code Evaluation
Using Monte Carlo Tree Search to improve LLM-based code assessment
AIDE: AI Revolutionizing ML Development
Automating the tedious trial-and-error in machine learning engineering
Enhancing Code Reasoning in LLMs
A new approach to improve how AI models understand and reason through code
Automating API Knowledge Graph Construction
A novel framework for rich, reliable API insights without manual schema design
Automating Docker Configurations with AI
How LLM-based agents can streamline development environments
Trustworthy AI Coding Assistants
Building reliable agentic AI systems for software engineering
Quality Over Quantity in AI Test Generation
How data quality dramatically improves automated unit testing
SolSearch: An LLM-Driven Framework for Efficient SAT-Solving...
By Junjie Sheng, Yanqiu Lin...
Democratizing Application Development
Bridging the Gap Between LLM Code Generation and Deployment
Benchmarking LLMs for GPU Optimization
Evaluating AI's ability to generate efficient Triton code for GPU kernels
Breaking the Text-to-SQL Barrier
A Multi-Module Framework for Enhanced Language-to-Query Conversion
Sol-Ver: Self-Improvement for Code Generation
Teaching AI to Write Better Code Through Self-Play and Verification
Building Better Software Verification
First comprehensive dataset of C++ formal specifications
Benchmarking LLMs for Smarter Code Completion
Evaluating modern AI models for context-aware programming assistance
Revolutionizing API Development with AI
Leveraging LLMs for Faster, More Efficient API Synthesis
Smarter Bug Fixing Through Hierarchical Localization
How LLMs are revolutionizing software debugging workflows
LLMs as Specification Engineers
Leveraging AI to Write Formal Software Specifications
Smarter Code Generation Through Pragmatic Reasoning
How LLMs can better understand ambiguous user instructions
LLMs & Code Review Accountability
How AI is reshaping developer responsibility in code reviews
Smarter Test Assertions with AI
Enhancing software testing through fine-tuned retrieval-augmented language models
Evaluating LLM Code Critique Capabilities
A new benchmark for assessing how well AI can analyze and improve code
Keeping LLMs in Sync with Evolving Code
A novel approach to updating AI models with current programming practices
Speeding Up AI Code Generation
A specialized approach to accelerate LLM inference for coding tasks
Enhancing LLM Reasoning for Software Engineering
Applying Reinforcement Learning to Software Evolution Tasks
AI-Powered Design Pattern Detection
Using Large Language Models to Improve Software Architecture Analysis
AI-Powered Bug Fixing Agents
Automating code debugging with LLMs and LangGraph
Optimizing Code Generation with LLMs
A framework to enhance both efficiency and correctness in AI-generated code
Beyond Tool-Based AI Programming
A New Paradigm for Software Engineering Agents Using Visual Interfaces
Benchmarking LLMs for AI Development
A new standard for evaluating code generation in deep learning workflows
Smarter Commit Messages with AI
Leveraging LLMs and In-Context Learning for Code Documentation
AI-Powered Class Invariant Generation
Using LLMs to Automate Critical Software Specifications
Unlocking LLM Code Generation Capabilities
Separating problem-solving from language-coding skills
Evaluating LLMs' Code Generation Abilities
First benchmark for measuring how well AI follows instructions when writing code
LLMs as Software Architects?
Evaluating AI's Capability to Understand Complex Software Design Patterns
The Future of Software Engineering in the AI Era
Why LLMs won't replace software engineers anytime soon
Consistency in AI Code Reviews
Evaluating how deterministic LLMs are when reviewing code
Automating GUI Design with AI
How LLMs can revolutionize software prototyping
Rethinking LLM Testing
A new taxonomic approach for testing language model software
Enhancing Bug Report Classification with Machine Learning
Improving Development Efficiency Through Automated Classification
Improving Developer Questions with AI
Using LLMs for Named Entity Recognition in Developer Conversations
Democratizing Code Generation with LLMs
Breaking barriers between natural language and programming
Bridging the Gap: Vision-Language Models for Front-End Development
Enhancing AI-powered code generation through targeted data synthesis
LLMs as Code Quality Judges
Automating software evaluation with AI
Engineering the Prompt Layer
Bringing Software Engineering Discipline to LLM Prompt Development
PolyVer: Revolutionizing Multi-Language Software Verification
A compositional approach to verify polyglot software systems
Protecting Software Architecture with AI
Enhancing accuracy in detecting architectural tactics using prompts
AI-Powered Debugging for Hardware Design
Using Large Language Models to Solve RTL Assertion Failures
Rethinking LLMs for Type Inference in Code
Evaluating true capabilities beyond memorization
Enhancing Code Repair with LLMs
Leveraging language models to extract fix ingredients beyond context windows
Testing LLMs' Code Error Prediction
A novel benchmark for evaluating code understanding beyond synthesis
Testing LLM Agent Tools Automatically
Ensuring reliable tool documentation for AI agents
Transforming Software Development with LLMs
How AI is reshaping developers, processes, and products
Testing LLM Prompts: The Missing Quality Assurance Layer
Automated testing for prompt engineering with PromptPex
Smarter AI-Powered Test Generation
Using Static Analysis to Enhance LLM-Generated Unit Tests
Teaching LLMs to Understand Code Execution
Training Models with Dynamic Program Behavior, Not Just Static Code
Decoding Developer Psychology
Using Psycholinguistic Analysis to Understand Software Engineering Teams
Optimizing LLM Code Generation
A Comprehensive Taxonomy of Inefficiencies in AI-Generated Code
Smarter Code Generation with LLMs
Improving code quality through Comparative Prefix-Tuning
Proactive Web Accessibility in Your IDE
LLM-powered VS Code extension for real-time accessibility improvements
Enhancing REST API Testing with LLMs
Leveraging ChatGPT and GitHub Copilot for stronger test automation
Illuminating the Black Box of ML Systems
A novel approach to analyze the semantic flow within ML components
DynaCode: Rethinking Code Generation Benchmarks
A Dynamic Approach to Combat LLM Memorization in Code Evaluation
Team of LLMs: Better Together
Leveraging Multi-Agent Collaboration for GitHub Documentation Summarization
Adapting LLMs to Engineer Problem-Solving Styles
Creating more inclusive AI assistants for software development
LLMs as Bug Replicators
How language models perpetuate coding errors during code completion
Automating Prompt Engineering for Better Code
AI-powered prompt refinement to enhance code generation capabilities
Enhancing Commit Messages with Human Context
Optimizing developer communication through LLM-assisted refinement
Improving Code Understanding with Multi-Agent Debate
How collaborative LLM agents enhance code summarization and translation
AI-Powered Hazard Analysis
Enhancing STPA with Large Language Models
UML Code Generation From Images
Using Multimodal LLMs to Transform Visual Diagrams to Executable Code
Why AI Code Agents Fail at GitHub Issue Resolution
Uncovering the pitfalls in LLM-based software development agents
Modular Prompting Revolution
Breaking the linear chain for better code generation
AI-Powered C to Rust Translation
Enhancing code safety with LLM-driven multi-step translation
Panta: Enhancing Test Generation with LLMs
Combining AI and program analysis for better code coverage
LLMs and Performance Optimization: A Critical Assessment
Evaluating LLMs' effectiveness in optimizing high-performance computing code
Trust in AI: The Double-Edged Sword
Building well-calibrated trust for LLMs in Software Engineering
Synthetic Data Revolution with LLMs
Leveraging AI to create high-quality training data for text and code
Enhancing AI Coding Assistants
A Dynamic Approach to Improve LLM Decision-Making in Coding Tasks
MANTRA: AI-Powered Code Refactoring
Automating method-level refactoring through multi-agent LLM collaboration
EnvBench: Automating Development Environment Setup
A benchmark for evaluating LLM capabilities in software configuration tasks
CodingGenie: AI That Anticipates Developer Needs
Transforming coding workflows with proactive LLM assistance
Community Knowledge Powers Better Code Reasoning
Leveraging Stack Overflow to enhance LLM reasoning capabilities
Conversational LLMs for Code Repair
Understanding the effectiveness and limitations of ChatGPT-based bug fixing
LogiAgent: Smart API Testing with AI
Multi-agent LLM framework detects logical bugs in REST APIs
Enhancing AI Code Generation with Human Feedback
A Bayesian approach to crowd-sourced reinforcement learning for better coding assistants
Supercharging LLMs with API Knowledge
How Retrieval Augmentation Boosts Code Generation Capabilities
Automating Non-Functional Requirements with AI
Using LLMs to enhance software quality from the start
Reinventing AI Code Generation
Integrating Software Engineering Principles into LLMs
Smarter Code Completion with Long-Context LLMs
Training models to effectively utilize repository-wide context
Smarter Code Generation with LLMs
Reducing 'overthinking' to improve efficiency & accuracy
Breaking Down Complex Programming Tasks
How LLMs Can Solve Programming Tasks by Composition
Boosting Code Generation with LLM Ensembles
Using similarity-based selection to improve AI code generation accuracy
PolyTest: Next-Gen Unit Testing with LLMs
Leveraging language diversity and self-consistency for more robust test generation
Bridging the Gap: LLMs and Code Review Comprehension
Evaluating how well AI understands human code feedback
Understanding Programmers' LLM Interactions
How developers use AI coding assistants and what they expect
Making LLMs Write Better Code
Fault-Aware Fine-Tuning for Improved Code Generation Accuracy
Evolving with Rust: LLMs and API Challenges
A dynamic benchmark for evaluating version-aware code generation
The Python Preference of LLMs
LLMs show significant bias in programming language selection
LLMs Revolutionizing Code Analysis
Transforming Software Engineering Through AI-Powered Code Understanding
Safer C to Rust Migration
AI-powered code transformation with memory safety guarantees
Enhancing Code Generation with Documentation
Automating E2E test creation using LLMs and product documentation
Automating Test Case Generation with AI
Using LLMs to bridge business requirements and software testing
Beyond Function-Level: Intelligent Code Translation
Triple Knowledge Augmentation for Repository-Context Translation
Enhancing AI Coding Assistants
How Verbal Process Supervision Creates Better Coding Agents
Benchmarking LLaMA2 for Code Development
Evaluating AI capabilities across programming languages for scientific applications
Boosting CPU Performance with AI
How LLMs transform code for better vectorization
Bridging Natural Language and Code Verification
Using structured reasoning to verify program correctness
Optimizing AI Coding Assistants for Developer Experience
SLA-Driven Architecture for Responsive CodeLLMs
Automating Software Architecture with AI
LLM-powered agents for intelligent architecture design
Enhancing Test Automation with AI
Leveraging LLMs and Case-Based Reasoning for Functional Test Script Generation
Harnessing LLMs for Bug Diagnostics
Using AI to extract failure-inducing inputs from bug reports
Optimizing Code Generation with Smart Retrieval
Identifying what context matters most for AI code assistants
Smarter Tool Use for AI Models
Improving LLM capabilities through code-based process supervision
AI-Powered Code Refactoring
Using LLMs to Automate Method Relocation in Software Engineering
Interactive Debugging for LLMs
Enabling AI to explore and debug code like humans
KGCompass: Smarter Software Repairs
Enhancing Bug Fixes Using Repository-Aware Knowledge Graphs
AI-Powered Automotive Software Validation
Using Reasoning-Enhanced LLM Agents for Precision Release Analytics
Smart Bug Hunting with AI
Using LLMs to efficiently locate software issues across codebases
Evaluating LLMs for Interactive Code Generation
New benchmark reveals how well AI follows instructions when writing code
Testing LLM Robustness for Software Requirements
Evaluating consistency in NFR-aware code generation
AI-Driven Root Cause Analysis
Solving Distributed System Failures with Code-Enhanced LLMs
Testing LLM Reasoning in Code Synthesis
Evaluating how AI agents learn from examples using a novel interactive benchmark
Smart Code Completion with Context
Enhancing LLMs for Complex Data Transfer Tasks
Onboarding Buddy: AI-Powered Developer Onboarding
Revolutionizing software engineering onboarding with LLMs and RAG
LLMigrate: Transforming "Lazy" Large Language Models into Ef...
By Yuchen Liu, Junhao Hu...
Rust Compiler Bugs: Hidden Challenges in Safe Systems
Uncovering unique bug patterns in Rust's safety enforcement mechanisms
Self-Rectifying Code Generation
Enabling Non-Coders to Build Complete Projects with AI
Supercharging Bug Detection with LLMs
Using Large Language Models to improve bug-inducing commit identification
Boosting Automated Program Repair with LLMs
Knowledge Prompt Tuning for Enhanced Bug Fixes with Limited Data
AI Copilot: Revolutionizing Software Testing
Bridging Code Generation and Bug Detection with Context-Based RAG
Unlocking Coding Capabilities in LLMs
Making LLM reasoning more accessible through data distillation
Making LLMs Better at Long Code Translation
Improving accuracy through instrumentation and state alignment
Human Factors in LLM Adoption
Understanding why engineers choose LLMs for specific tasks
Expanding LLM Evaluation Beyond Python
A new multilingual benchmark for code issue resolution
Unlocking Code-Capable AI
How OpenCodeInstruct Advances Code Generation in LLMs
Enhancing LLM Code Generation
Automating Prompt Engineering with Diffusion Models
Evaluating LLMs' Code Comprehension
Measuring how well AI models truly understand code
Dynamic LLM Code Evaluation Reimagined
Beyond Static Benchmarks with Monte Carlo Tree Search
Enhancing API Testing with LLM Intelligence
Overcoming Testing Plateaus with AI-Assisted Mutation
ARLO: Automating Architecture from Requirements
Using LLMs to transform natural language requirements into software architecture
Scaling EDA Testing with Synthetic Verilog
Generating realistic HDL designs to improve tool robustness and LLM training
Enhancing LLMs for Code Repair
New benchmark evaluates how well AI models use feedback to fix bugs
Automating Acceptance Testing with LLMs
Transforming User Stories into Executable Test Scripts
Line-by-Line Code Generation
A novel approach to improve LLM code quality and efficiency
Cross-Domain Code Search Without Retraining
A Novel Contrastive Learning Approach for Zero-Shot Retrieval
Supercharging REST API Testing with AI
How LLM-based systems amplify test coverage and bug detection
Automating Architecture Decisions with AI
Leveraging LLMs to streamline software architecture documentation
AI-Powered System Integration
How LLMs can automate enterprise service composition
Evaluating Code Assistants: Beyond the Hype
A robust methodology for measuring AI coding tool quality
AI-Powered User Story Generation
Automating Requirements for Software Product Lines with LLMs
Benchmarking Coding Assistants Across Languages
A new multi-language benchmark for evaluating AI coding agents in real-world scenarios
DocAgent: Revolutionizing Code Documentation
A multi-agent system that enhances automated documentation quality and accuracy
Smarter Compiler Testing with AI
Using RAG to automate and enhance cross-architecture compiler verification
Smarter Code Generation with Type Constraints
Reducing compilation errors in AI-generated code through type-aware LLMs
Automating Linux Kernel Patch Migration with AI
Using Large Language Models to reduce manual engineering effort
Automating Code Migration with LLMs
How Google scales software evolution using AI
Enhancing LLM Code Generation
Making libraries more accessible to AI through specialized documentation
CodeRAG: Enhancing Real-World Code Generation
A bigraph-based retrieval system for complex programming environments
Multilingual Code Generation Breakthrough
Enhancing programming with structured reasoning across languages
Automating COBOL to Java Validation
AI-powered testing framework for legacy code transformation
Enhancing Software Requirements Quality
A conceptual framework to overcome natural language ambiguity
AI-Powered Engineering Assistance
Enhancing Test Data Analytics with LLM-Based Reasoning
Building a Standard Library for LLMs
Standardizing APIs for Retrieval-Augmented Generation
Swarm Agent: Revolutionizing MLOps
Enhancing Human-Machine Collaboration through Conversational AI
