EvoOptiGraph: Weakness-Driven Coevolution via Graph-Based Structural Generation for Optimization Modeling
Summary
EvoOptiGraph, a novel framework, addresses two key challenges in automating optimization modeling from natural language with large language models (LLMs): the lack of structural diversity in training corpora and static data generation pipelines. This system represents each mixed-integer linear program (MILP) as an attributed bipartite graph, employing validity-preserving evolutionary operators to generate structurally diverse instances. These evolved graphs are then converted into solver code and natural language. Training proceeds in two stages: supervised fine-tuning (SFT) on an initial dataset, followed by reinforcement learning with verifiable rewards (RLVR), where graph-derived weakness signals guide the generation of new instances targeting the model's failures. This closed-loop coevolution of data and model significantly outperforms larger generalist models, agentic methods, and specialized baselines in accuracy, executability, and generalization across six public datasets.
Key takeaway
For AI Scientists and Machine Learning Engineers developing LLMs for optimization modeling, you should consider implementing data-model coevolution strategies. EvoOptiGraph demonstrates that guiding data generation with model weaknesses significantly boosts accuracy and generalization. Integrate graph-based structural generation and a two-stage SFT/RLVR training approach to overcome limitations of static training corpora and achieve superior performance on MILP tasks.
Key insights
Targeted data-model coevolution, driven by model weaknesses, significantly improves LLMs for optimization modeling.
Principles
- Structural diversity in training data is crucial for LLM generalization.
- Weakness-driven data generation enhances model performance.
- Representing MILPs as graphs enables validity-preserving evolution.
Method
EvoOptiGraph represents MILPs as attributed bipartite graphs, applies evolutionary operators for diverse instance generation, then converts graphs to solver code and natural language for two-stage SFT and RLVR training.
In practice
- Use graph-based representations for structural data generation.
- Implement weakness-driven feedback loops in LLM training.
- Employ coevolutionary strategies for data and model improvement.
Topics
- Optimization Modeling
- Large Language Models
- Data-Model Coevolution
- Mixed-Integer Linear Programs
- Graph-Based Generation
- Reinforcement Learning
Best for: Research Scientist, AI Scientist, Machine Learning Engineer
Related on AIssential
Editorial summary, takeaway, and curation by AIssential. Original article published by Artificial Intelligence.