Bridging Literature and the Universe Via A Multi-Agent Large Language Model System

2025-07-11

AI Summary

Cosmological simulation parameter extraction suffers from heterogeneous literature formats, error-prone and inefficient manual conversion. To address this, we propose SimAgents—the first multi-agent large language model system tailored for astrophysics—integrating domain-specific physical reasoning, tool-augmented execution, and structured inter-agent communication to enable automated parameter extraction, cross-document consistency verification, and generation of executable simulation scripts. We introduce a novel benchmark dataset comprising over 40 real-world cosmological simulations and publicly release both the system and dataset. Experiments demonstrate that SimAgents significantly outperforms baseline methods in parameter extraction accuracy and script syntactic/semantic compliance. By bridging the gap between scientific literature and numerical simulation infrastructure, SimAgents enhances research efficiency and reproducibility in computational cosmology.

Technology Category

• Multiagent Systems: Agent-Based Simulation and Emergent Behavior
• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Agentic search
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems

Abstract

As cosmological simulations and their associated software become increasingly complex, physicists face the challenge of searching through vast amounts of literature and user manuals to extract simulation parameters from dense academic papers, each using different models and formats. Translating these parameters into executable scripts remains a time-consuming and error-prone process. To improve efficiency in physics research and accelerate the cosmological simulation process, we introduce SimAgents, a multi-agent system designed to automate both parameter configuration from the literature and preliminary analysis for cosmology research. SimAgents is powered by specialized LLM agents capable of physics reasoning, simulation software validation, and tool execution. These agents collaborate through structured communication, ensuring that extracted parameters are physically meaningful, internally consistent, and software-compliant. We also construct a cosmological parameter extraction evaluation dataset by collecting over 40 simulations in published papers from Arxiv and leading journals that cover diverse simulation types. Experiments on the dataset demonstrate a strong performance of SimAgents, highlighting its effectiveness and potential to accelerate scientific research for physicists. Our demonstration video is available at: https://youtu.be/w1zLpm_CaWA. The complete system and dataset are publicly available at https://github.com/xwzhang98/SimAgents.

Problem

• Automate extraction of simulation parameters from dense literature
• Translate parameters into executable scripts efficiently
• Ensure parameters are physically meaningful and software-compliant

Innovation

• Multi-agent LLM system automates parameter extraction
• Agents ensure physics validation and software compliance
• Public dataset with 40+ simulations for evaluation