
Leveraging the multilingual capabilities of large language models (LLMs), InterTrans improves automated code translation by strategically using intermediate programming languages as stepping stones between the source and target languages.


coremsg

intertrans-enhancing-large-language-model-based-code-translation-through-transitive-intermediate-translations


InterTrans: Enhancing Large Language Model-Based Code Translation Through Transitive Intermediate Translations


title_rewrite


과학 컴퓨팅에서 레거시 포트란 코드를 C++로 변환하는 데 대형 언어 모델(LLM) 기반 도구인 [Code-Scribe]를 활용하면 개발 생산성을 향상시킬 수 있다.


대형-언어-모델을-활용한-과학-컴퓨팅-분야의-코드-변환-및-소프트웨어-개발-mcfm-코드-변환-사례-연구-및-code-scribe-도구-소개


대형 언어 모델을 활용한 과학 컴퓨팅 분야의 코드 변환 및 소프트웨어 개발: MCFM 코드 변환 사례 연구 및 [Code-Scribe] 도구 소개



Data augmentation techniques, specifically using comparable corpora and generating multiple reference translations, can significantly improve the performance of code translation models, especially when parallel training data is scarce.


improving-code-translation-with-limited-data-leveraging-comparable-corpora-and-multiple-references


Improving Code Translation with Limited Data: Leveraging Comparable Corpora and Multiple References
