This research addresses critical limitations in how Large Language Models translate complex, lengthy code between programming languages.

Key Innovations:

• Uses program instrumentation to capture execution behavior
• Aligns program states between source and target code
• Significantly improves translation accuracy for longer, more complex code
• Enables better preservation of functionality across languages

Engineering Impact: This approach helps software teams migrate codebases between languages more reliably, reducing errors during cross-platform development and modernization projects.

Enhancing LLMs in Long Code Translation through Instrumentation and Program State Alignment