Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Cai, Junyang, Kadioglu, Serdar, Dilkina, Bistra
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2509.08970
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866916986755219456
author Cai, Junyang
Kadioglu, Serdar
Dilkina, Bistra
author_facet Cai, Junyang
Kadioglu, Serdar
Dilkina, Bistra
contents Natural language descriptions of optimization or satisfaction problems are challenging to translate into correct MiniZinc models, as this process demands both logical reasoning and constraint programming expertise. We introduce Gala, a framework that addresses this challenge with a global agentic approach: multiple specialized large language model (LLM) agents decompose the modeling task by global constraint type. Each agent is dedicated to detecting and generating code for a specific class of global constraint, while a final assembler agent integrates these constraint snippets into a complete MiniZinc model. By dividing the problem into smaller, well-defined sub-tasks, each LLM handles a simpler reasoning challenge, potentially reducing overall complexity. We conduct initial experiments with several LLMs and show better performance against baselines such as one-shot prompting and chain-of-thought prompting. Finally, we outline a comprehensive roadmap for future work, highlighting potential enhancements and directions for improvement.
format Preprint
id arxiv_https___arxiv_org_abs_2509_08970
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Gala: Global LLM Agents for Text-to-Model Translation
Cai, Junyang
Kadioglu, Serdar
Dilkina, Bistra
Artificial Intelligence
Natural language descriptions of optimization or satisfaction problems are challenging to translate into correct MiniZinc models, as this process demands both logical reasoning and constraint programming expertise. We introduce Gala, a framework that addresses this challenge with a global agentic approach: multiple specialized large language model (LLM) agents decompose the modeling task by global constraint type. Each agent is dedicated to detecting and generating code for a specific class of global constraint, while a final assembler agent integrates these constraint snippets into a complete MiniZinc model. By dividing the problem into smaller, well-defined sub-tasks, each LLM handles a simpler reasoning challenge, potentially reducing overall complexity. We conduct initial experiments with several LLMs and show better performance against baselines such as one-shot prompting and chain-of-thought prompting. Finally, we outline a comprehensive roadmap for future work, highlighting potential enhancements and directions for improvement.
title Gala: Global LLM Agents for Text-to-Model Translation
topic Artificial Intelligence
url https://arxiv.org/abs/2509.08970