Saved in:
Bibliographic Details
Main Authors: Zhang, Mian, Liu, Shujian, Dong, Sixun, Yin, Ming, Hu, Yebowen, Wang, Xun, Ma, Steven, Wang, Song, Indurthi, Sathish Reddy, Deng, Haoyun, Chen, Zhiyu Zoey, Song, Kaiqiang
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2508.09125
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911400592408576
author Zhang, Mian
Liu, Shujian
Dong, Sixun
Yin, Ming
Hu, Yebowen
Wang, Xun
Ma, Steven
Wang, Song
Indurthi, Sathish Reddy
Deng, Haoyun
Chen, Zhiyu Zoey
Song, Kaiqiang
author_facet Zhang, Mian
Liu, Shujian
Dong, Sixun
Yin, Ming
Hu, Yebowen
Wang, Xun
Ma, Steven
Wang, Song
Indurthi, Sathish Reddy
Deng, Haoyun
Chen, Zhiyu Zoey
Song, Kaiqiang
contents Instruction following has catalyzed the recent era of Large Language Models (LLMs) and is the foundational skill underpinning more advanced capabilities such as reasoning and agentic behaviors. As tasks grow more challenging, the logic structures embedded in natural language instructions becomes increasingly intricate. However, how well LLMs perform on such logic-rich instructions remains under-explored. We propose LogicIFGen and LogicIFEval. LogicIFGen is a scalable, automated framework for generating verifiable instructions from code functions, which can naturally express rich logic such as conditions, loops, and function calls. We further curate a collection of complex code functions and use LogicIFGen to construct LogicIFEval, a benchmark comprising 426 verifiable logic-rich instructions. Our experiments demonstrate that current state-of-the-art LLMs still struggle to correctly follow the instructions in LogicIFEval. Most LLMs can only follow fewer than 60% of the instructions, revealing significant deficiencies in the instruction-following ability. Code and Benchmark: https://github.com/mianzhang/LogicIF
format Preprint
id arxiv_https___arxiv_org_abs_2508_09125
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Complex Logical Instruction Generation
Zhang, Mian
Liu, Shujian
Dong, Sixun
Yin, Ming
Hu, Yebowen
Wang, Xun
Ma, Steven
Wang, Song
Indurthi, Sathish Reddy
Deng, Haoyun
Chen, Zhiyu Zoey
Song, Kaiqiang
Computation and Language
Machine Learning
Instruction following has catalyzed the recent era of Large Language Models (LLMs) and is the foundational skill underpinning more advanced capabilities such as reasoning and agentic behaviors. As tasks grow more challenging, the logic structures embedded in natural language instructions becomes increasingly intricate. However, how well LLMs perform on such logic-rich instructions remains under-explored. We propose LogicIFGen and LogicIFEval. LogicIFGen is a scalable, automated framework for generating verifiable instructions from code functions, which can naturally express rich logic such as conditions, loops, and function calls. We further curate a collection of complex code functions and use LogicIFGen to construct LogicIFEval, a benchmark comprising 426 verifiable logic-rich instructions. Our experiments demonstrate that current state-of-the-art LLMs still struggle to correctly follow the instructions in LogicIFEval. Most LLMs can only follow fewer than 60% of the instructions, revealing significant deficiencies in the instruction-following ability. Code and Benchmark: https://github.com/mianzhang/LogicIF
title Complex Logical Instruction Generation
topic Computation and Language
Machine Learning
url https://arxiv.org/abs/2508.09125