Abstract:Large language models have transformed AI-assisted software engineering, but current research remains biased toward high-resource languages such as Python, with weaker performance in languages like Rust and OCaml. Since real-world systems are inherently polyglot, robust multilingual code intelligence is crucial. This survey focuses on two key tasks: multilingual code generation from shared natural-language requirements, and multilingual code translation that preserves semantics across languages. It reviews representative methods, benchmarks, and evaluation metrics, and highlights challenges and opportunities for trustworthy cross-language generalization.
| Subjects: | Software Engineering (cs.SE); Machine Learning (cs.LG); Programming Languages (cs.PL) |
| Cite as: | arXiv:2604.25960 [cs.SE] |
| (or arXiv:2604.25960v1 [cs.SE] for this version) | |
| https://doi.org/10.48550/arXiv.2604.25960 arXiv-issued DOI via DataCite |
Submission history
From: Cheng Wen [view email]
[v1]
Mon, 27 Apr 2026 20:20:26 UTC (4,258 KB)