Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2510.05228 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915822568472576 |
|---|---|
| author | Pan, Haining Roggeveen, James V. Berg, Erez Carrasquilla, Juan Chowdhury, Debanjan Ganguli, Surya Ghimenti, Federico Hasik, Juraj Hunt, Henry Jiang, Hong-Chen Kamb, Mason Kao, Ying-Jer Khatami, Ehsan Lawler, Michael J. Luo, Di Neupert, Titus Qi, Xiaoliang Brenner, Michael P. Kim, Eun-Ah |
| author_facet | Pan, Haining Roggeveen, James V. Berg, Erez Carrasquilla, Juan Chowdhury, Debanjan Ganguli, Surya Ghimenti, Federico Hasik, Juraj Hunt, Henry Jiang, Hong-Chen Kamb, Mason Kao, Ying-Jer Khatami, Ehsan Lawler, Michael J. Luo, Di Neupert, Titus Qi, Xiaoliang Brenner, Michael P. Kim, Eun-Ah |
| contents | Large language models (LLMs) have shown remarkable progress in coding and math problem-solving, but evaluation on advanced research-level problems in hard sciences remains scarce. To fill this gap, we present CMT-Benchmark, a dataset of 50 problems covering condensed matter theory (CMT) at the level of an expert researcher. Topics span analytical and computational approaches in quantum many-body, and classical statistical mechanics. The dataset was designed and verified by a panel of expert researchers from around the world. We built the dataset through a collaborative environment that challenges the panel to write and refine problems they would want a research assistant to solve, including Hartree-Fock, exact diagonalization, quantum/variational Monte Carlo, density matrix renormalization group (DMRG), quantum/classical statistical mechanics, and model building. We evaluate LLMs by programmatically checking solutions against expert-supplied ground truth. We developed machine-grading, including symbolic handling of non-commuting operators via normal ordering. They generalize across tasks too. Our evaluations show that frontier models struggle with all of the problems in the dataset, highlighting a gap in the physical reasoning skills of current LLMs. Notably, experts identified strategies for creating increasingly difficult problems by interacting with the LLMs and exploiting common failure modes. The best model, GPT5, solves 30\% of the problems; average across 17 models (GPT, Gemini, Claude, DeepSeek, Llama) is 11.4\pm2.1\%. Moreover, 18 problems are solved by none of the 17 models, and 26 by at most one. These unsolved problems span Quantum Monte Carlo, Variational Monte Carlo, and DMRG. Answers sometimes violate fundamental symmetries or have unphysical scaling dimensions. We believe this benchmark will guide development toward capable AI research assistants and tutors. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_05228 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | CMT-Benchmark: A Benchmark for Condensed Matter Theory Built by Expert Researchers Pan, Haining Roggeveen, James V. Berg, Erez Carrasquilla, Juan Chowdhury, Debanjan Ganguli, Surya Ghimenti, Federico Hasik, Juraj Hunt, Henry Jiang, Hong-Chen Kamb, Mason Kao, Ying-Jer Khatami, Ehsan Lawler, Michael J. Luo, Di Neupert, Titus Qi, Xiaoliang Brenner, Michael P. Kim, Eun-Ah Machine Learning Artificial Intelligence Large language models (LLMs) have shown remarkable progress in coding and math problem-solving, but evaluation on advanced research-level problems in hard sciences remains scarce. To fill this gap, we present CMT-Benchmark, a dataset of 50 problems covering condensed matter theory (CMT) at the level of an expert researcher. Topics span analytical and computational approaches in quantum many-body, and classical statistical mechanics. The dataset was designed and verified by a panel of expert researchers from around the world. We built the dataset through a collaborative environment that challenges the panel to write and refine problems they would want a research assistant to solve, including Hartree-Fock, exact diagonalization, quantum/variational Monte Carlo, density matrix renormalization group (DMRG), quantum/classical statistical mechanics, and model building. We evaluate LLMs by programmatically checking solutions against expert-supplied ground truth. We developed machine-grading, including symbolic handling of non-commuting operators via normal ordering. They generalize across tasks too. Our evaluations show that frontier models struggle with all of the problems in the dataset, highlighting a gap in the physical reasoning skills of current LLMs. Notably, experts identified strategies for creating increasingly difficult problems by interacting with the LLMs and exploiting common failure modes. The best model, GPT5, solves 30\% of the problems; average across 17 models (GPT, Gemini, Claude, DeepSeek, Llama) is 11.4\pm2.1\%. Moreover, 18 problems are solved by none of the 17 models, and 26 by at most one. These unsolved problems span Quantum Monte Carlo, Variational Monte Carlo, and DMRG. Answers sometimes violate fundamental symmetries or have unphysical scaling dimensions. We believe this benchmark will guide development toward capable AI research assistants and tutors. |
| title | CMT-Benchmark: A Benchmark for Condensed Matter Theory Built by Expert Researchers |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2510.05228 |