Leveraging Code Language Models for Code Smells Detection

CodeLite: A Low-Cost Framework for Code Classification Tasks via Enhanced Code Representations and Lexical Fusion

The prevailing approach in software engineering for code classification tasks is to rely on large-scale code language models such as StarCoder and CodeLlama. While these models have demonstrated strong performance, they come with the price of intensive computations that raise energy concerns and limit their deployment in resource-constrained environments. In this work, we propose ‘CodeLite’, a relatively lightweight framework that integrates moderately sized code models, such as CodeBERT, with traditional frequency-based techniques, achieving better performance and computational efficiency compared to larger models. Our approach comprises multiple stages, starting with the enhancement of encoder-based code models, where we leverage their intermediate-layer information to capture richer lexical and syntactic signals beyond the default [CLS] token. In parallel, we employ a regex-enhanced TF-IDF component tailored for source code to capture complementary frequency-based patterns. These representations are then combined with the learned code embeddings through suitable fusion techniques for final decision-making. Evaluated across multiple downstream tasks, including programming language identification, authorship attribution, and AI plagiarism, CodeLite consistently outperforms strong baseline models, achieving up to 4.2% absolute accuracy improvement over other approaches and 11.8% incremental gain over ablation components while incurring significantly lower training costs. We further conduct several statistical tests and ablation studies to validate the contribution of each component in the proposed framework. Overall, our results demonstrate that a careful combination of lightweight models along with task-aware feature enhancements can serve as a practical and efficient alternative to heavyweight language models for various software engineering tasks.

An Exploratory Study on Code Smells Detection and Refactoring using LLMs

This paper focuses on four common smells and considers five prompts of diverse complexity, asking the model for detecting and removing the identified code smells, and suggests that general-purpose LLMs cannot be reliably used for that.

Enhancing code smell classification with code refactoring-based data augmentation

Machine Learning for Code Smell Detection and Resolution

Code smells are indicative of poor software design, maintainability issues, or potential defects, and their early detection is critical for high-quality software development. Traditional detection methods rely heavily on manual inspection or rule-based static analysis, which are often time-consuming, error-prone, and limited in adaptability. This paper explores the use of machine learning techniques for automated code smell detection and resolution. By leveraging code metrics, syntactic and semantic features, and historical refactoring data, ML models can identify patterns associated with common code smells and recommend targeted resolution strategies. The proposed approach is evaluated on open-source and industrial software projects, demonstrating improvements in detection accuracy, maintainability, and developer productivity. The study highlights the potential of intelligent, data-driven approaches to enhance software quality and support continuous code improvement.

Multilingual Source Code Vulnerability Detection Using Deep Learning: A Semantic Representation and Transfer Learning Approach

Detecting vulnerabilities in source code remains a major challenge as modern software systems increasingly span multiple programming languages. This work presents a deep learning approach for multilingual vulnerability detection that emphasizes semantic transfer rather than architectural complexity. The proposed method combines contextualized embeddings from CodeBERT/GraphCodeBERT with a BiLSTM and attention mechanism to capture code semantics and adopts a cross-lingual transfer setting where models trained on one language (e.g., Python) are evaluated on another (e.g., Java). To improve robustness under data imbalance, SMOTE and stratified cross-validation are incorporated into the training process. Experiments on Juliet and CodeXGLUE show that the model achieves an F1-score of about 0.87 and a ROC-AUC of 0.85 in intra-language settings, while guided fine-tuning improves cross-language F1-score by an average of 0.18 and ROC-AUC by approximately 0.13 compared with the direct transfer baseline. These results suggest that stabilizing semantic representations during transfer is key to improving generalization while maintaining practical efficiency under moderate computational constraints.

Selecting and Combining Large Language Models in Scalable Code Clone Detection

Findings indicate that ensembling approaches can be statistically significant and effective on larger datasets, where the best-performing ensemble improved performance by 37% over its individual LLMs on the commercial large-scale code.