Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Huang, Yue, Jiang, Zhengzhe, Luo, Xiaonan, Guo, Kehan, Zhuang, Haomin, Zhou, Yujun, Yuan, Zhengqing, Sun, Xiaoqi, Schleinitz, Jules, Wang, Yanbo, Zhang, Shuhao, Surve, Mihir, Chawla, Nitesh V, Wiest, Olaf, Zhang, Xiangliang
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2509.16543
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866911165671538688
author Huang, Yue
Jiang, Zhengzhe
Luo, Xiaonan
Guo, Kehan
Zhuang, Haomin
Zhou, Yujun
Yuan, Zhengqing
Sun, Xiaoqi
Schleinitz, Jules
Wang, Yanbo
Zhang, Shuhao
Surve, Mihir
Chawla, Nitesh V
Wiest, Olaf
Zhang, Xiangliang
author_facet Huang, Yue
Jiang, Zhengzhe
Luo, Xiaonan
Guo, Kehan
Zhuang, Haomin
Zhou, Yujun
Yuan, Zhengqing
Sun, Xiaoqi
Schleinitz, Jules
Wang, Yanbo
Zhang, Shuhao
Surve, Mihir
Chawla, Nitesh V
Wiest, Olaf
Zhang, Xiangliang
contents Empowering large language models (LLMs) with chemical intelligence remains a challenge due to the scarcity of high-quality, domain-specific instruction-response datasets and the misalignment of existing synthetic data generation pipelines with the inherently hierarchical and rule-governed structure of chemical information. To address this, we propose ChemOrch, a framework that synthesizes chemically grounded instruction-response pairs through a two-stage process: task-controlled instruction generation and tool-aware response construction. ChemOrch enables controllable diversity and levels of difficulty for the generated tasks, and ensures response precision through tool planning and distillation, and tool-based self-repair mechanisms. The effectiveness of ChemOrch is evaluated based on: 1) the high quality of generated instruction data, demonstrating superior diversity and strong alignment with chemical constraints; 2) the reliable generation of evaluation tasks that more effectively reveal LLM weaknesses in chemistry; and 3) the significant improvement of LLM chemistry capabilities when the generated instruction data are used for fine-tuning. Our work thus represents a critical step toward scalable and verifiable chemical intelligence in LLMs.
format Preprint
id arxiv_https___arxiv_org_abs_2509_16543
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle ChemOrch: Empowering LLMs with Chemical Intelligence via Synthetic Instructions
Huang, Yue
Jiang, Zhengzhe
Luo, Xiaonan
Guo, Kehan
Zhuang, Haomin
Zhou, Yujun
Yuan, Zhengqing
Sun, Xiaoqi
Schleinitz, Jules
Wang, Yanbo
Zhang, Shuhao
Surve, Mihir
Chawla, Nitesh V
Wiest, Olaf
Zhang, Xiangliang
Computation and Language
Empowering large language models (LLMs) with chemical intelligence remains a challenge due to the scarcity of high-quality, domain-specific instruction-response datasets and the misalignment of existing synthetic data generation pipelines with the inherently hierarchical and rule-governed structure of chemical information. To address this, we propose ChemOrch, a framework that synthesizes chemically grounded instruction-response pairs through a two-stage process: task-controlled instruction generation and tool-aware response construction. ChemOrch enables controllable diversity and levels of difficulty for the generated tasks, and ensures response precision through tool planning and distillation, and tool-based self-repair mechanisms. The effectiveness of ChemOrch is evaluated based on: 1) the high quality of generated instruction data, demonstrating superior diversity and strong alignment with chemical constraints; 2) the reliable generation of evaluation tasks that more effectively reveal LLM weaknesses in chemistry; and 3) the significant improvement of LLM chemistry capabilities when the generated instruction data are used for fine-tuning. Our work thus represents a critical step toward scalable and verifiable chemical intelligence in LLMs.
title ChemOrch: Empowering LLMs with Chemical Intelligence via Synthetic Instructions
topic Computation and Language
url https://arxiv.org/abs/2509.16543