InverseCoder: Self-improving Instruction-Tuned Code LLMs with Inverse-Instruct

Recent advancements in open-source code large language models (LLMs) have been driven by fine-tuning on the data generated from powerful closed-source LLMs, which are expensive to obtain. This paper explores whether it is possible to use a fine-tuned open-source model to generate additional data to augment its instruction-tuning dataset. We make two observations: (1) A code snippet can serve as the response to different instructions. (2) Instruction-tuned code LLMs perform better at translating code into instructions than the reverse. Based on these observations, we propose Inverse-Instruct, a data augmentation technique that uses a fine-tuned LLM to generate additional instructions of code responses from its own training dataset. The additional instruction-response pairs are added to the original dataset, and a stronger code LLM can be obtained by fine-tuning on the augmented dataset. We empirically validate Inverse-Instruct on a range of open-source code models (e.g. CodeLlama-Python and DeepSeek-Coder) and benchmarks (e.g., HumanEval(+), MBPP(+), DS-1000 and MultiPL-E), showing it consistently improves the base models.

Authors

Institutions

Publication Details

Journal
Proceedings of the AAAI Conference on Artificial Intelligence
Published
2025-04-11
DOI
https://doi.org/10.1609/aaai.v39i24.34742
Primary Topic
Microfluidic and Capillary Electrophoresis Applications
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

InverseCoder: Self-improving Instruction-Tuned Code LLMs with Inverse-Instruct

Xishan Zhang, Wei Wang, S. B. Liu, Ziyuan Nan et al.
Proceedings of the AAAI Conference on Artificial Intelligence
Microfluidic and Capillary Electrophoresis Applications
article

InverseCoder: Self-improving Instruction-Tuned Code LLMs with Inverse-Instruct

Xishan Zhang, Wei Wang, S. B. Liu, Ziyuan Nan, Yunji Chen, Kaizhao Yuan, Yewen Pu, Zidong Du, Yutong Wu, Rui Zhang, Xing Jun Hu, Lingzhe Gao, Di Huang, Qi Guo, Wenxuan Shi, Dawei Yin
article en

Abstract

Recent advancements in open-source code large language models (LLMs) have been driven by fine-tuning on the data generated from powerful closed-source LLMs, which are expensive to obtain. This paper explores whether it is possible to use a fine-tuned open-source model to generate additional data to augment its instruction-tuning dataset. We make two observations: (1) A code snippet can serve as the response to different instructions. (2) Instruction-tuned code LLMs perform better at translating code into instructions than the reverse. Based on these observations, we propose Inverse-Instruct, a data augmentation technique that uses a fine-tuned LLM to generate additional instructions of code responses from its own training dataset. The additional instruction-response pairs are added to the original dataset, and a stronger code LLM can be obtained by fine-tuning on the augmented dataset. We empirically validate Inverse-Instruct on a range of open-source code models (e.g. CodeLlama-Python and DeepSeek-Coder) and benchmarks (e.g., HumanEval(+), MBPP(+), DS-1000 and MultiPL-E), showing it consistently improves the base models.

Proceedings of the AAAI Conference on Artificial IntelligenceVol. 39(24)
Autodesk (United States) (US), Institute of Computing Technology (CN), Baidu (China) (CN)
National Science Foundation, National Natural Science Foundation of China, Chinese Academy of Sciences, Youth Innovation Promotion Association of the Chinese Academy of Sciences, Youth Innovation Promotion Association
Openalex Percentile: Top 90%
Microfluidic and Capillary Electrophoresis Applications
DOIOpen Access PDFOpenAlex
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.