Eliciting Instruction-tuned Code Language Models' Capabilities to Utilize Auxiliary Function for Code Generation

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Lee, Seonghyeon, Kim, Suyeon, Jang, Joonwon, Chon, Heejae, Lee, Dongha, Yu, Hwanjo
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866916405574631424
author Lee, Seonghyeon
Kim, Suyeon
Jang, Joonwon
Chon, Heejae
Lee, Dongha
Yu, Hwanjo
author_facet Lee, Seonghyeon
Kim, Suyeon
Jang, Joonwon
Chon, Heejae
Lee, Dongha
Yu, Hwanjo
contents We study the code generation behavior of instruction-tuned models built on top of code pre-trained language models when they could access an auxiliary function to implement a function. We design several ways to provide auxiliary functions to the models by adding them to the query or providing a response prefix to incorporate the ability to utilize auxiliary functions with the instruction-following capability. Our experimental results show the effectiveness of combining the base models' auxiliary function utilization ability with the instruction following ability. In particular, the performance of adopting our approaches with the open-sourced language models surpasses that of the recent powerful proprietary language models, i.e., gpt-4o.
format Preprint
id arxiv_https___arxiv_org_abs_2409_13928
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Eliciting Instruction-tuned Code Language Models' Capabilities to Utilize Auxiliary Function for Code Generation
Lee, Seonghyeon
Kim, Suyeon
Jang, Joonwon
Chon, Heejae
Lee, Dongha
Yu, Hwanjo
Software Engineering
Artificial Intelligence
Computation and Language
We study the code generation behavior of instruction-tuned models built on top of code pre-trained language models when they could access an auxiliary function to implement a function. We design several ways to provide auxiliary functions to the models by adding them to the query or providing a response prefix to incorporate the ability to utilize auxiliary functions with the instruction-following capability. Our experimental results show the effectiveness of combining the base models' auxiliary function utilization ability with the instruction following ability. In particular, the performance of adopting our approaches with the open-sourced language models surpasses that of the recent powerful proprietary language models, i.e., gpt-4o.
title Eliciting Instruction-tuned Code Language Models' Capabilities to Utilize Auxiliary Function for Code Generation
topic Software Engineering
Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2409.13928