TriFusion-LLM: Prior-Guided Multimodal Fusion with LLM Arbitration for Fine-grained Code Clone Detection

2026-03-16

AI Summary

This work addresses the limitation of existing code clone detection methods, which are predominantly confined to binary classification and struggle to distinguish among the seven fine-grained clone types in BigCloneBench. To overcome this challenge, we propose a prior-guided multimodal fusion framework that jointly integrates heuristic similarity priors, abstract syntax tree structural features, and CodeBERT semantic embeddings. Furthermore, we introduce a large language model (LLM)-based selective arbitration mechanism that performs inference augmentation only on high-uncertainty samples—approximately 0.2% of the total—thereby maintaining low computational overhead. Our approach achieves a significant improvement in Macro-F1 score on the seven-class BigCloneBench task, raising it from 0.695 to 0.878, thus effectively balancing performance and efficiency.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Code clone detection (CCD) supports software maintenance, refactoring, and security analysis. Although pre-trained models capture code semantics, most work reduces CCD to binary classification, overlooking the heterogeneity of clone types and the seven fine-grained categories in BigCloneBench. We present Full Model, a multimodal fusion framework that jointly integrates heuristic similarity priors from classical machine learning, structural signals from abstract syntax trees (ASTs), and deep semantic embeddings from CodeBERT into a single predictor. By fusing structural, statistical, and semantic representations, Full Model improves discrimination among fine-grained clone types while keeping inference cost practical. On the seven-class BigCloneBench benchmark, Full Model raises Macro-F1 from 0.695 to 0.875. Ablation studies show that using the primary model's probability distribution as a prior to guide selective arbitration by a large language model (LLM) substantially outperforms blind reclassification; arbitrating only ~0.2% of high-uncertainty samples yields an additional 0.3 absolute Macro-F1 gain. Overall, Full Model achieves an effective performance-cost trade-off for fine-grained CCD and offers a practical solution for large-scale industrial deployment.

Problem

• code clone detection
• fine-grained classification
• multimodal fusion
• BigCloneBench
• clone type heterogeneity

Innovation

• multimodal fusion
• LLM arbitration
• code clone detection
• fine-grained classification
• prior-guided learning