Structured Chemistry Reasoning with Large Language Models

Siru Ouyang, Zhuosheng Zhang, Bing Yan, Xuan Liu, Yejin Choi, Jiawei Han, Lianhui Qin
Proceedings of the 41st International Conference on Machine Learning, PMLR 235:38937-38952, 2024.

Abstract

Large Language Models (LLMs) excel in diverse areas, yet struggle with complex scientific reasoning, especially in the field of chemistry. Different from the simple chemistry tasks (e.g., molecule classification) addressed in previous studies, complex chemistry problems require not only vast knowledge and precise calculation, but also compositional reasoning about rich dynamic interactions of different concepts (e.g., temperature changes). Our study shows that even advanced LLMs, like GPT-4, can fail easily in different ways. Interestingly, the errors often stem not from a lack of domain knowledge within the LLMs, but rather from the absence of an effective reasoning structure that guides the LLMs to elicit the right knowledge, incorporate the knowledge in step-by-step reasoning, and iteratively refine results for further improved quality. On this basis, we introduce StructChem, a simple yet effective prompting strategy that offers the desired guidance and substantially boosts the LLMs’ chemical reasoning capability. Testing across four chemistry areas—quantum chemistry, mechanics, physical chemistry, and kinetics—StructChem substantially enhances GPT-4’s performance, with up to 30% peak improvement. Our analysis also underscores the unique difficulties of precise grounded reasoning in science with LLMs, highlighting a need for more research in this area.

Cite this Paper

BibTeX
@InProceedings{pmlr-v235-ouyang24a, title = {Structured Chemistry Reasoning with Large Language Models}, author = {Ouyang, Siru and Zhang, Zhuosheng and Yan, Bing and Liu, Xuan and Choi, Yejin and Han, Jiawei and Qin, Lianhui}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, pages = {38937--38952}, year = {2024}, editor = {Salakhutdinov, Ruslan and Kolter, Zico and Heller, Katherine and Weller, Adrian and Oliver, Nuria and Scarlett, Jonathan and Berkenkamp, Felix}, volume = {235}, series = {Proceedings of Machine Learning Research}, month = {21--27 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v235/main/assets/ouyang24a/ouyang24a.pdf}, url = {https://proceedings.mlr.press/v235/ouyang24a.html}, abstract = {Large Language Models (LLMs) excel in diverse areas, yet struggle with complex scientific reasoning, especially in the field of chemistry. Different from the simple chemistry tasks (e.g., molecule classification) addressed in previous studies, complex chemistry problems require not only vast knowledge and precise calculation, but also compositional reasoning about rich dynamic interactions of different concepts (e.g., temperature changes). Our study shows that even advanced LLMs, like GPT-4, can fail easily in different ways. Interestingly, the errors often stem not from a lack of domain knowledge within the LLMs, but rather from the absence of an effective reasoning structure that guides the LLMs to elicit the right knowledge, incorporate the knowledge in step-by-step reasoning, and iteratively refine results for further improved quality. On this basis, we introduce StructChem, a simple yet effective prompting strategy that offers the desired guidance and substantially boosts the LLMs’ chemical reasoning capability. Testing across four chemistry areas—quantum chemistry, mechanics, physical chemistry, and kinetics—StructChem substantially enhances GPT-4’s performance, with up to 30% peak improvement. Our analysis also underscores the unique difficulties of precise grounded reasoning in science with LLMs, highlighting a need for more research in this area.} }
Endnote
%0 Conference Paper %T Structured Chemistry Reasoning with Large Language Models %A Siru Ouyang %A Zhuosheng Zhang %A Bing Yan %A Xuan Liu %A Yejin Choi %A Jiawei Han %A Lianhui Qin %B Proceedings of the 41st International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2024 %E Ruslan Salakhutdinov %E Zico Kolter %E Katherine Heller %E Adrian Weller %E Nuria Oliver %E Jonathan Scarlett %E Felix Berkenkamp %F pmlr-v235-ouyang24a %I PMLR %P 38937--38952 %U https://proceedings.mlr.press/v235/ouyang24a.html %V 235 %X Large Language Models (LLMs) excel in diverse areas, yet struggle with complex scientific reasoning, especially in the field of chemistry. Different from the simple chemistry tasks (e.g., molecule classification) addressed in previous studies, complex chemistry problems require not only vast knowledge and precise calculation, but also compositional reasoning about rich dynamic interactions of different concepts (e.g., temperature changes). Our study shows that even advanced LLMs, like GPT-4, can fail easily in different ways. Interestingly, the errors often stem not from a lack of domain knowledge within the LLMs, but rather from the absence of an effective reasoning structure that guides the LLMs to elicit the right knowledge, incorporate the knowledge in step-by-step reasoning, and iteratively refine results for further improved quality. On this basis, we introduce StructChem, a simple yet effective prompting strategy that offers the desired guidance and substantially boosts the LLMs’ chemical reasoning capability. Testing across four chemistry areas—quantum chemistry, mechanics, physical chemistry, and kinetics—StructChem substantially enhances GPT-4’s performance, with up to 30% peak improvement. Our analysis also underscores the unique difficulties of precise grounded reasoning in science with LLMs, highlighting a need for more research in this area.
APA
Ouyang, S., Zhang, Z., Yan, B., Liu, X., Choi, Y., Han, J. & Qin, L.. (2024). Structured Chemistry Reasoning with Large Language Models. Proceedings of the 41st International Conference on Machine Learning, in Proceedings of Machine Learning Research 235:38937-38952 Available from https://proceedings.mlr.press/v235/ouyang24a.html.

Related Material