CodeChemist: Functional Knowledge Transfer for Low-Resource Code Generation via Test-Time Scaling

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, Kaixin, Li, Tianlin, Zhang, Xiaoyu, Liu, Aishan, Liu, Xianglong, Liu, Ziqi, Zhang, Zhiqiang, Zhou, Jun, Shi, and Bin
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908571654946816
author Wang, Kaixin
Li, Tianlin
Zhang, Xiaoyu
Liu, Aishan
Liu, Xianglong
Liu, Ziqi
Zhang, Zhiqiang
Zhou, Jun
Shi, and Bin
author_facet Wang, Kaixin
Li, Tianlin
Zhang, Xiaoyu
Liu, Aishan
Liu, Xianglong
Liu, Ziqi
Zhang, Zhiqiang
Zhou, Jun
Shi, and Bin
contents Code Large Language Models (CodeLLMs) are increasingly used in code generation tasks across a wide range of applications. However, their performance is often inconsistent across different programming languages (PLs), with low-resource PLs suffering the most due to limited training data. In this paper, we present CodeChemist, a novel and efficient framework for test-time scaling that enables functional knowledge transfer from high-resource to low-resource PLs using generated test cases. CodeChemist first generates and executes code in high-resource PLs to create test cases that encapsulate functional knowledge. It then uses multi-temperature hedged sampling to generate code snippets in the low-resource PL and selects the best one based on the pass rate of the test cases. Our extensive experiments show that CodeChemist outperforms existing test-time scaling approaches, boosting the performance of code generation for low-resource PLs without requiring any model retraining.
format Preprint
id arxiv_https___arxiv_org_abs_2510_00501
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle CodeChemist: Functional Knowledge Transfer for Low-Resource Code Generation via Test-Time Scaling
Wang, Kaixin
Li, Tianlin
Zhang, Xiaoyu
Liu, Aishan
Liu, Xianglong
Liu, Ziqi
Zhang, Zhiqiang
Zhou, Jun
Shi, and Bin
Software Engineering
Code Large Language Models (CodeLLMs) are increasingly used in code generation tasks across a wide range of applications. However, their performance is often inconsistent across different programming languages (PLs), with low-resource PLs suffering the most due to limited training data. In this paper, we present CodeChemist, a novel and efficient framework for test-time scaling that enables functional knowledge transfer from high-resource to low-resource PLs using generated test cases. CodeChemist first generates and executes code in high-resource PLs to create test cases that encapsulate functional knowledge. It then uses multi-temperature hedged sampling to generate code snippets in the low-resource PL and selects the best one based on the pass rate of the test cases. Our extensive experiments show that CodeChemist outperforms existing test-time scaling approaches, boosting the performance of code generation for low-resource PLs without requiring any model retraining.
title CodeChemist: Functional Knowledge Transfer for Low-Resource Code Generation via Test-Time Scaling
topic Software Engineering
url https://arxiv.org/abs/2510.00501