ChemDFM-R: A Chemical Reasoning LLM Enhanced with Atomized Chemical Knowledge
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866915935523176448 |
|---|---|
| author | Zhao, Zihan Wan, Ziping Chen, Lu Lin, Xuanze Yu, Shiyang Zhang, Situo Ma, Da Zhu, Zichen Zhang, Danyang Wang, Huayang Dai, Zhongyang Wen, Liyang Chen, Bo Chen, Xin Yu, Kai |
| author_facet | Zhao, Zihan Wan, Ziping Chen, Lu Lin, Xuanze Yu, Shiyang Zhang, Situo Ma, Da Zhu, Zichen Zhang, Danyang Wang, Huayang Dai, Zhongyang Wen, Liyang Chen, Bo Chen, Xin Yu, Kai |
| contents | Atomized chemical knowledge, such as functional group information of molecules and reactions, plays a pivotal intermediate role in the reasoning process that connects molecular structures with their properties and reactivities. While large language models (LLMs) have achieved impressive progress, the absence of atomized chemical knowledge results in their superficial understanding of chemistry and limited chemical reasoning capabilities. In this work, to tackle this problem, we develop a Chemical Reasoning LLM, ChemDFM-R. We first construct a comprehensive dataset of atomized chemical knowledge, ChemFG, annotating the presence of functional groups in molecules and the changes of functional groups during chemical reactions, to enhance the model's understanding of the fundamental principles and internal logic of chemistry. Then, we propose a mixed-source distillation method that initializes the model's reasoning capability with limited distilled data, and develop a four-stage training pipeline to equip the model with atomized chemical knowledge and chemical reasoning logic. Experiments on diverse chemical benchmarks demonstrate that ChemDFM-R achieves cutting-edge performance while providing interpretable, rationale-driven outputs, surpassing both the general-domain LLMs and domain-specific chemical LLMs. Moreover, ChemDFM-R achieves comparable or superior performance compared with cutting-edge commercial LLMs, such as o4-mini. Further case studies illustrate how explicit reasoning chains significantly improve the model's reliability, transparency, and practicality in real-world human-AI collaboration scenarios. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_21990 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | ChemDFM-R: A Chemical Reasoning LLM Enhanced with Atomized Chemical Knowledge Zhao, Zihan Wan, Ziping Chen, Lu Lin, Xuanze Yu, Shiyang Zhang, Situo Ma, Da Zhu, Zichen Zhang, Danyang Wang, Huayang Dai, Zhongyang Wen, Liyang Chen, Bo Chen, Xin Yu, Kai Computational Engineering, Finance, and Science Artificial Intelligence Atomized chemical knowledge, such as functional group information of molecules and reactions, plays a pivotal intermediate role in the reasoning process that connects molecular structures with their properties and reactivities. While large language models (LLMs) have achieved impressive progress, the absence of atomized chemical knowledge results in their superficial understanding of chemistry and limited chemical reasoning capabilities. In this work, to tackle this problem, we develop a Chemical Reasoning LLM, ChemDFM-R. We first construct a comprehensive dataset of atomized chemical knowledge, ChemFG, annotating the presence of functional groups in molecules and the changes of functional groups during chemical reactions, to enhance the model's understanding of the fundamental principles and internal logic of chemistry. Then, we propose a mixed-source distillation method that initializes the model's reasoning capability with limited distilled data, and develop a four-stage training pipeline to equip the model with atomized chemical knowledge and chemical reasoning logic. Experiments on diverse chemical benchmarks demonstrate that ChemDFM-R achieves cutting-edge performance while providing interpretable, rationale-driven outputs, surpassing both the general-domain LLMs and domain-specific chemical LLMs. Moreover, ChemDFM-R achieves comparable or superior performance compared with cutting-edge commercial LLMs, such as o4-mini. Further case studies illustrate how explicit reasoning chains significantly improve the model's reliability, transparency, and practicality in real-world human-AI collaboration scenarios. |
| title | ChemDFM-R: A Chemical Reasoning LLM Enhanced with Atomized Chemical Knowledge |
| topic | Computational Engineering, Finance, and Science Artificial Intelligence |
| url | https://arxiv.org/abs/2507.21990 |