Developers use LLMs like ChatGPT mainly for knowledge acquisition and code generation at the detailed design level, reporting benefits such as better technology selection and early flaw detection alongside limitations like lengthy outputs, incorrect code, and hallucinations.
Title resolution pending
19 Pith papers cite this work. Polarity classification is still indexing.
19
Pith papers citing it
citation-role summary
background 2
dataset 2
citation-polarity summary
representative citing papers
Speculative decoding accelerates LLM inference on SE tasks without accuracy loss, with model-based methods suiting code generation and model-free methods suiting repository-level repair and editing.
ADI equips AI debugging agents with function-level interaction via a new execution trace structure, raising SWE-bench Verified resolution to 63.8% at $1.28 per task and delivering 6-18% gains when added to existing agents.
GenAI use in undergraduate software projects leads to four patterns of Comprehension Debt accumulation in students' code understanding, with one pattern where AI instead scaffolds deeper learning.
Multiple-choice queries synthesized from Hoare triples enable more reliable identification of intended programs than labeled-example supervision in active learning for program disambiguation.
LLMs display clear performance stratification on formal language tasks aligned with Chomsky hierarchy complexity levels, limited by severe efficiency barriers rather than absolute capability.
Build-bench is the first architecture-aware benchmark that evaluates LLMs on repairing cross-ISA build failures via iterative tool-augmented reasoning, with the best model reaching 63.19% success.
Introduces contextualized code pretraining with caller-callee pairs from static analysis to train CallerGen models that outperform baselines on the new CallerEval benchmark.
A review of 114 studies creates taxonomies for code and data quality issues, formalizes 18 propagation mechanisms from training data defects to LLM-generated code defects, and synthesizes detection and mitigation techniques.
REC RL improves LLM code generation by automatically assessing and optimizing requirement difficulty with adaptive curriculum sampling, yielding 1.23-5.62% Pass@1 gains over baselines.
SpecPylot generates and validates icontract specifications for Python programs by combining LLM proposals with Crosshair symbolic execution feedback.
LLM-powered monitoring of UI similarity allows random testing tools to escape tarpits, yielding 45-55% higher coverage and more unique bugs across 12 apps.
Reconstructing 6946 syzbot bug-fix lifecycles reveals that accepted kernel patches are non-local and reviewer-constrained, enabling PatchAdvisor to improve automated repair quality over baselines via retrieval and diagnostic guidance.
By proving test suite coverage is monotone submodular and training LLMs with RL to maximize marginal gains, TestDecision improves branch coverage 38-52% and bug detection up to 95% over base models on ULT and LiveCodeBench.
RSA prompting enables LLMs to automatically create functional exploits for CVEs in Odoo ERP, succeeding on all tested cases in 3-5 rounds and removing the need for manual effort.
LLM-generated code matches human-written code in overall readability but exhibits different issue patterns, and prompt engineering has limited impact on improving it.
Fine-tuning and prompting reduce some CWEs in AI-generated code but frequently introduce new weaknesses, with no strategy working reliably across models or languages.
MultiVul uses multimodal contrastive learning to align code and comment representations, yielding up to 27% F1 gains on vulnerability detection benchmarks over prompting and code-only baselines.
Auto-Diagnose applies LLMs to summarize and diagnose root causes of integration test failures, reporting 90.14% accuracy on 71 manual cases and positive adoption after Google-wide rollout.
citing papers explorer
-
Using LLMs in Software Design: An Empirical Study of GitHub and A Practitioner Survey
Developers use LLMs like ChatGPT mainly for knowledge acquisition and code generation at the detailed design level, reporting benefits such as better technology selection and early flaw detection alongside limitations like lengthy outputs, incorrect code, and hallucinations.
-
An Empirical Study of Speculative Decoding on Software Engineering Tasks
Speculative decoding accelerates LLM inference on SE tasks without accuracy loss, with model-based methods suiting code generation and model-free methods suiting repository-level repair and editing.
-
Empowering Autonomous Debugging Agents with Efficient Dynamic Analysis
ADI equips AI debugging agents with function-level interaction via a new execution trace structure, raising SWE-bench Verified resolution to 63.8% at $1.28 per task and delivering 6-18% gains when added to existing agents.
-
Comprehension Debt in GenAI-Assisted Software Engineering Projects
GenAI use in undergraduate software projects leads to four patterns of Comprehension Debt accumulation in students' code understanding, with one pattern where AI instead scaffolds deeper learning.
-
Choose, Don't Label: Multiple-Choice Query Synthesis for Program Disambiguation
Multiple-choice queries synthesized from Hoare triples enable more reliable identification of intended programs than labeled-example supervision in active learning for program disambiguation.
-
Evaluating the Formal Reasoning Capabilities of Large Language Models through Chomsky Hierarchy
LLMs display clear performance stratification on formal language tasks aligned with Chomsky hierarchy complexity levels, limited by severe efficiency barriers rather than absolute capability.
-
Can Language Models Go Beyond Coding? Assessing the Capability of Language Models to Build Real-World Systems
Build-bench is the first architecture-aware benchmark that evaluates LLMs on repairing cross-ISA build failures via iterative tool-augmented reasoning, with the best model reaching 63.19% success.
-
Contextualized Code Pretraining for Code Generation
Introduces contextualized code pretraining with caller-callee pairs from static analysis to train CallerGen models that outperform baselines on the new CallerEval benchmark.
-
Bridging Generation and Training: A Systematic Review of Quality Issues in LLMs for Code
A review of 114 studies creates taxonomies for code and data quality issues, formalizes 18 propagation mechanisms from training data defects to LLM-generated code defects, and synthesizes detection and mitigation techniques.
-
Improving LLM Code Generation via Requirement-Aware Curriculum Reinforcement Learning
REC RL improves LLM code generation by automatically assessing and optimizing requirement difficulty with adaptive curriculum sampling, yielding 1.23-5.62% Pass@1 gains over baselines.
-
SpecPylot: Python Specification Generation using Large Language Models
SpecPylot generates and validates icontract specifications for Python programs by combining LLM proposals with Crosshair symbolic execution feedback.
-
Improving Random Testing via LLM-powered UI Tarpit Escaping for Mobile Apps
LLM-powered monitoring of UI similarity allows random testing tools to escape tarpits, yielding 45-55% higher coverage and more unique bugs across 12 apps.
-
Beyond Crash-to-Patch: Patch Evolution for Linux Kernel Repair
Reconstructing 6946 syzbot bug-fix lifecycles reveals that accepted kernel patches are non-local and reviewer-constrained, enabling PatchAdvisor to improve automated repair quality over baselines via retrieval and diagnostic guidance.
-
TestDecision: Sequential Test Suite Generation via Greedy Optimization and Reinforcement Learning
By proving test suite coverage is monotone submodular and training LLMs with RL to maximize marginal gains, TestDecision improves branch coverage 38-52% and bug detection up to 95% over base models on ULT and LiveCodeBench.
-
From Rookie to Expert: Manipulating LLMs for Automated Vulnerability Exploitation in Enterprise Software
RSA prompting enables LLMs to automatically create functional exploits for CVEs in Odoo ERP, succeeding on all tested cases in 3-5 rounds and removing the need for manual effort.
-
The Readability Spectrum: Patterns, Issues, and Prompt Effects in LLM-Generated Code
LLM-generated code matches human-written code in overall readability but exhibits different issue patterns, and prompt engineering has limited impact on improving it.
-
On Fixing Insecure AI-Generated Code through Model Fine-Tuning and Prompting Strategies
Fine-tuning and prompting reduce some CWEs in AI-generated code but frequently introduce new weaknesses, with no strategy working reliably across models or languages.
-
Learning Generalizable Multimodal Representations for Software Vulnerability Detection
MultiVul uses multimodal contrastive learning to align code and comment representations, yielding up to 27% F1 gains on vulnerability detection benchmarks over prompting and code-only baselines.
-
LLM-Based Automated Diagnosis Of Integration Test Failures At Google
Auto-Diagnose applies LLMs to summarize and diagnose root causes of integration test failures, reporting 90.14% accuracy on 71 manual cases and positive adoption after Google-wide rollout.