UrbanMind: Towards Urban General Intelligence via Tool-Enhanced Retrieval-Augmented Generation and Multilevel Optimization

2025/07/07 by Yang, Kai, Zhu, Zelin, Jian, Chengtao +4
#Artificial Intelligence (cs.AI) #FOS: Computer and information sciences #Machine Learning (cs.LG)

paper · doi:10.48550/arxiv.2507.04706

Abstract

Urban general intelligence (UGI) refers to the capacity of AI systems to autonomously perceive, reason, and act within dynamic and complex urban environments. In this paper, we introduce UrbanMind, a tool-enhanced retrieval-augmented generation (RAG) framework designed to facilitate UGI. Central to UrbanMind is a novel architecture based on Continual Retrieval-Augmented MoE-based LLM (C-RAG-LLM), which dynamically incorporates domain-specific knowledge and evolving urban data to support long-term adaptability. The architecture of C-RAG-LLM aligns naturally with a multilevel optimization framework, where different layers are treated as interdependent sub-problems. Each layer has distinct objectives and can be optimized either independently or jointly through a hierarchical learning process. The framework is highly flexible, supporting both end-to-end training and partial layer-wise optimization based on resource or deployment constraints. To remain adaptive under data drift, it is further integrated with an incremental corpus updating mechanism. Evaluations on real-world urban tasks of a variety of complexity verify the effectiveness of the proposed framework. This work presents a promising step toward the realization of general-purpose LLM agents in future urban environments.

Citations

• A Survey of Personalization: From RAG to Agent
• Adaptive Robust Optimization Models for DER Planning in Distribution Networks under Long- and Short-Term Uncertainties
• Anomaly Detection in Event-triggered Traffic Time Series via Similarity Learning
• RAG-Enhanced Collaborative LLM Agents for Drug Discovery
• Research on the Online Update Method for Retrieval-Augmented Generation (RAG) Model with Incremental Learning
• Reinforcement Learning Enhanced LLMs: A Survey
• A Stochastic Approach to Bi-Level Optimization for Hyperparameter Optimization and Meta Learning
• Tri-Level Navigator: LLM-Empowered Tri-Level Learning for Time Series OOD Generalization
• Retrieval Augmented Generation (RAG) and Beyond: A Comprehensive Survey on How to Make your LLMs use External Data More Wisely
• Evaluation of RAG Metrics for Question Answering in the Telecom Domain
• AvaTaR: Optimizing LLM Agents for Tool Usage via Contrastive Reasoning
• DR-RAG: Applying Dynamic Document Relevance to Retrieval-Augmented Generation for Question-Answering
• A Review of Prominent Paradigms for LLM-Based Agents: Tool Use (Including RAG), Planning, and Feedback Learning
• Multi-level Personalized Federated Learning on Heterogeneous and Long-Tailed Data
• Continual Learning for Smart City: A Survey
• MetaSTNet: Multimodal Meta-learning for Cellular Traffic Conformal Prediction
• RealGen: Retrieval Augmented Generation for Controllable Traffic Scenarios
• Levels of AGI for Operationalizing Progress on the Path to AGI
• Llama 2: Open Foundation and Fine-Tuned Chat Models
• TA-MoE: Topology-Aware Large Scale Mixture-of-Expert Training
• Asynchronous Distributed Bilevel Optimization
• MegaBlocks: Efficient Sparse Training with Mixture-of-Experts
• Cellular Traffic Prediction via Deep State Space Models with Attention Mechanism
• From Concept Drift to Model Degradation: An Overview on Performance-Aware Drift Detectors
• GLaM: Efficient Scaling of Language Models with Mixture-of-Experts
• Recent Advances of Continual Learning in Computer Vision: An Overview
• A Bilevel Programming Framework for Joint Edge Resource Management and Pricing
• TrafficStream: A Streaming Traffic Flow Forecasting Framework Based on Graph Neural Networks and Continual Learning
• Investigating Bi-Level Optimization for Learning and Vision from a Unified Perspective: A Survey and Beyond
• GPT-3: Its Nature, Scope, Limits, and Consequences
• Efficient Architecture Search for Continual Learning
• Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
• Toward Understanding Catastrophic Forgetting in Continual Learning
• Robust Optimization over Multiple Domains
• City Foundation Models for Learning General Purpose Representations from OpenStreetMap
• AGI: Artificial General Intelligence for Education
• BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

Related

• Urban-R1: Reinforced MLLMs Mitigate Geospatial Biases for Urban General Intelligence
• Large Language Model Powered Intelligent Urban Agents: Concepts, Capabilities, and Applications
• Urban Generative Intelligence (UGI): A Foundational Platform for Agents in Embodied City Environment
• Urban-MAS: Human-Centered Urban Prediction with LLM-Based Multi-Agent System
• Towards Urban General Intelligence: A Review and Outlook of Urban Foundation Models