The paper introduces OptiMUS, a Large Language Model (LLM)-based agent designed to formulate and solve linear programming problems from their natural language descriptions. OptiMUS can develop mathematical models, write and debug solver code, and evaluate the generated solutions. The system outperforms existing methods on both easy and hard datasets. The paper also highlights the challenges that remain before LLMs can reliably model real-life optimization problems, such as dealing with ambiguous terms and long complex problem descriptions.
Publication date: 16 Feb 2024
Project Page: Not provided
Paper: https://arxiv.org/pdf/2402.10172
