OptiTree: Hierarchical Thoughts Generation with Tree Search for LLM Optimization Modeling
Haoyang Liu, Jie Wang, Yuyang Cai, Xiongwei Han, Yufei Kuang, Jianye Hao · 2025
Read OptiTree: Hierarchical Thoughts Generation with Tree Search for LLM Optimization Modeling by Haoyang Liu, Jie Wang, Yuyang Cai, Xiongwei Han, Yufei Kuang, Jianye Hao (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-22192.txt.
Description
Haoyang Liu, Jie Wang, Yuyang Cai, Xiongwei Han, Yufei Kuang, and Jianye Hao automate a hard translation in operations research. A wordy real problem must become variables, constraints, and an objective a solver can chew. Language models can draft such models. Fixed step-by-step prompting buckles on tangled mathematical structure, though.
OptiTree organizes the territory itself as a tree. Problem categories sit at the nodes, arranged by type and complexity. Each node is stocked with high-level modeling thoughts. Faced with a new problem, the system searches the tree repeatedly. The target breaks into a chain of simpler subproblems with known modeling patterns. The pieces are then composed. The point is adaptive decomposition, not a rigid recipe. Complex models get built the way a good analyst sketches them.
Catalog record
| CREATOR | Haoyang Liu, Jie Wang, Yuyang Cai, Xiongwei Han, Yufei Kuang, Jianye Hao |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 159 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-22192