BIM-Edit benchmark finds best LLM scores only 49.5% average across geometric, semantic, and topological metrics on 324 IFC editing tasks, with no model fully solving more than 3.4%.
hub
A Survey on Evaluating Large Language Models in Code Generation Tasks
14 Pith papers cite this work, alongside 9 external citations. Polarity classification is still indexing.
14
Pith papers citing it
9
external citations · Pith
abstract
This paper provides a comprehensive review of the current methods and metrics used to evaluate the performance of Large Language Models (LLMs) in code generation tasks. With the rapid growth in demand for automated software development, LLMs have demonstrated significant potential in the field of code generation. The paper begins by reviewing the historical development of LLMs and their applications in code generation. Next, it details various methods and metrics for assessing the code generation capabilities of LLMs, including code correctness, efficiency, readability, and evaluation methods based on expert review and user experience. The paper also evaluates the widely used benchmark datasets, identifying their limitations and proposing directions for future improvements. Specifically, the paper analyzes the performance of code generation models across different tasks by combining multiple evaluation metrics, such as code compilation/interpretation success rates, unit test pass rates, and performance and efficiency metrics, to comprehensively assess the practical application of LLMs in code generation. Finally, the paper discusses the challenges faced in evaluating LLMs in code generation, particularly how to ensure the comprehensiveness and accuracy of evaluation methods and how to adapt to the evolving practices of software development. These analyses and discussions provide valuable insights for further optimizing and improving the application of LLMs in code generation tasks.
hub tools
citation-role summary
background 3
dataset 1
citation-polarity summary
representative citing papers
A study of seven LLMs finds that realistic prompt variations such as one-character misspellings trigger library hallucinations in up to 26% of cases, fabricated names in up to 99%, and time-based prompts in up to 85%, and introduces LibHalluBench for evaluation.
On 2,520 programming tasks, matched Qwen general and coder models reliably raise Bloom cognitive demand but fail to lower it, so execution skill does not imply educational control.
TRAILS infers code correctness by aggregating LLM judgments on input-output pairs from category-partitioned specification tests, improving MCC by up to 39% over Zero-Shot COT on LiveCodeBench and CoCoClaNeL.
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.
Fine-tuning 7B code LLMs on a custom multi-file DSL dataset achieves structural fidelity of 1.00, high exact-match accuracy, and practical utility validated by expert survey and execution checks.
LLMs produce executable code only 42.55% of the time under API evolution without full documentation, improving to 66.36% with structured docs and by 11% more with reasoning strategies, yet outdated patterns persist.
A new benchmark, CLI-Tool-Bench, shows the best LLM agents achieve under ~43% semantic match when generating complete CLI tools from scratch.
Empirical study of eight LLMs finds overuse of popular libraries like NumPy in up to 45% of unnecessary cases and strong default preference for Python even when suboptimal.
LLM-generated cryptographic Rust code compiles successfully only 23% of the time and contains detectable vulnerabilities in 57% of the cases that do compile.
LLM-based SE tools lack stable ground truth and deterministic outputs, making standard evaluation assumptions invalid and requiring new approaches for reliable assessment.
Framework uses eight LLMs for code documentation generation and four LLMs as judges on nine criteria, showing 42% performance gap on medical physics library.
A systematic review finds research on the sustainability of LLM-generated code to be limited, fragmented, and without accepted frameworks for measurement or benchmarking.
citing papers explorer
-
BIM-Edit: Benchmarking Large Language Models for IFC-Based Building Information Modeling
BIM-Edit benchmark finds best LLM scores only 49.5% average across geometric, semantic, and topological metrics on 324 IFC editing tasks, with no model fully solving more than 3.4%.
-
Library Hallucinations in LLM-Generated Code: A Risk Analysis Grounded in Developer Queries
A study of seven LLMs finds that realistic prompt variations such as one-character misspellings trigger library hallucinations in up to 26% of cases, fabricated names in up to 99%, and time-based prompts in up to 85%, and introduces LibHalluBench for evaluation.
-
From Execution to Education: A Bloom-Aligned Framework for Measuring Educational Control in LLMs
On 2,520 programming tasks, matched Qwen general and coder models reliably raise Bloom cognitive demand but fail to lower it, so execution skill does not imply educational control.
-
Inferring Code Correctness from Specification
TRAILS infers code correctness by aggregating LLM judgments on input-output pairs from category-partitioned specification tests, improving MCC by up to 39% over Zero-Shot COT on LiveCodeBench and CoCoClaNeL.
-
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.
-
Leveraging LLMs for Multi-File DSL Code Generation: An Industrial Case Study
Fine-tuning 7B code LLMs on a custom multi-file DSL dataset achieves structural fidelity of 1.00, high exact-match accuracy, and practical utility validated by expert survey and execution checks.
-
When LLMs Lag Behind: Knowledge Conflicts from Evolving APIs in Code Generation
LLMs produce executable code only 42.55% of the time under API evolution without full documentation, improving to 66.36% with structured docs and by 11% more with reasoning strategies, yet outdated patterns persist.
-
Evaluating LLM-Based 0-to-1 Software Generation in End-to-End CLI Tool Scenarios
A new benchmark, CLI-Tool-Bench, shows the best LLM agents achieve under ~43% semantic match when generating complete CLI tools from scratch.
-
A Study of LLMs' Preferences for Libraries and Programming Languages
Empirical study of eight LLMs finds overuse of popular libraries like NumPy in up to 45% of unnecessary cases and strong default preference for Python even when suboptimal.
-
An Empirical Security Evaluation of LLM-Generated Cryptographic Rust Code
LLM-generated cryptographic Rust code compiles successfully only 23% of the time and contains detectable vulnerabilities in 57% of the cases that do compile.
-
Evaluation of LLM-Based Software Engineering Tools: Practices, Challenges, and Future Directions
LLM-based SE tools lack stable ground truth and deterministic outputs, making standard evaluation assumptions invalid and requiring new approaches for reliable assessment.
-
LLM-Based Code Documentation Generation and Multi-Judge Evaluation
Framework uses eight LLMs for code documentation generation and four LLMs as judges on nine criteria, showing 42% performance gap on medical physics library.
-
Sustainable Code Generation Using Large Language Models: A Systematic Literature Review
A systematic review finds research on the sustainability of LLM-generated code to be limited, fragmented, and without accepted frameworks for measurement or benchmarking.