Improving Code Translation with Syntax-Guided and Semantic-aware Preference Optimization
For developers and researchers using LLMs for code translation, this work addresses the bottleneck of unreliable semantic rewards by deriving robust signals directly from source code.
CTO improves LLM-based code translation by combining syntax-guided and semantic-aware preference optimization, achieving significant gains over baselines across C++, Java, and Python translations.
LLMs have shown immense potential for code translation, yet they often struggle to ensure both syntactic correctness and semantic consistency. While preference-based learning offers a promising alignment strategy, it is hindered by unreliable semantic rewards derived from sparse test cases or restrictive reference translations. We argue that a robust semantic reward for code translation must be derived directly from the source code. In this paper, we propose CTO to improve code translation with syntax-guided and semantic-aware preference optimization. Through contrastive learning, we train a cross-lingual semantic model to directly assess functional equivalence between source and translated code. By formulating code translation as a multi-objective optimization problem, this robust semantic signal is seamlessly unified with compiler-based syntactic feedback within the direct preference optimization framework. Extensive experiments on C++, Java, and Python translations demonstrate that CTO significantly outperforms existing baselines and alternative preference optimization strategies.