| Date | Name | Author | Institute | Links | Published |
|---|---|---|---|---|---|
| 2025-05 | MM-Agent: LLM as Agents for Real-World Mathematical Modeling Problem | Fan Liu, Zherui Yang, et al. | HKUST, et al. |   |
NeurIPS 2025 |
| 2025-05 | ModelingAgent: Bridging LLMs and Mathematical Modeling for Real-World Challenges | Cheng Qian, Hongyi Du, et al. | UIUC, et al. |   |
EMNLP (Findings) 2025 |
| 2025-05 | MathModelAgent | Jihe | - |  |
Code Release |
| 2024-05 | LLMs for Mathematical Modeling: Towards Bridging the Gap between Natural and Mathematical Languages | Xuhan Huang, Qingning Shen, et al. | CUHK(Shenzhen), et al. |   |
NAACL (Findings) 2025 |
| Coming Soon | LightningMathModel: Boosting Lightweight LLMs for Mathematical Modeling via Specialized Knowledge Training and Integration | - | 2026-01 |  |
| Name | Paper / Project | Description |
|---|---|---|
| MMBench | MM-Agent | A benchmark for evaluating LLMs' capabilities in real-world mathematical modeling, covering problem formulation, variable definition, and solution analysis. |
| ModelingBench | ModelingAgent | A benchmark designed to assess LLMs' performance across the complete mathematical modeling pipeline, from problem understanding to model construction and evaluation. |
| Mamo | Mamo | A benchmark focusing on the gap between natural language problem descriptions and formal mathematical language, evaluating LLMs' ability in model formulation and symbolic reasoning. |
| Name | Paper / Project | Links | Description |
|---|---|---|---|
| Mathematical_Modeling_Speciale_Dataset_v0.1 | LightningMathModel (IDEA Research) |  |
A specialized dataset for training and evaluating LLMs on mathematical modeling tasks, focusing on domain knowledge integration. |