Edge General Intelligence Through World Models and Agentic AI: Fundamentals, Solutions, and Challenges

arXiv 25.08 2025 58.9 survey

TLDR

A survey on integrating world models with agentic AI for edge general intelligence, covering foundations, applications, and challenges.

Reasoning

Strengths: Comprehensive survey bridging world models and edge AI, with clear applications. Weaknesses: Lacks empirical evaluation or real-world experiments; primarily conceptual.

Read-first score

Read-first score 58.9, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 49.

Recency 8%

86.7

Uses a gentle age decay so recent papers surface without erasing older foundations. 2025

Topical relevance 42%

Uses existing LLM keyword relevance scores normalized to 0-100. world model,world simulator,generative world model,interactive world model,video world model,world dynamics prediction,model-based reinforcement learning world model

Methodology quality 25%

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=analysis

Reproducibility 25%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=False; dataset=False; markers=none

Field roles

Frontier

Rank sensitivity

Stability: volatile; rank range: 365.

Keyword Scores

world model

world simulator

world dynamics prediction

generative world model

interactive world model

model-based reinforcement learning world model

video world model

Deep Analysis

Innovations

Bridging the gap between world models and wireless edge computing for Edge General Intelligence (EGI)
Proactive applications of world models in vehicular networks, UAV networks, IoT systems, and network functions virtualization
Synergy of world models with foundation models and digital twins as the cognitive backbone of EGI
Comprehensive analysis of architectural foundations including latent representation learning, dynamics modeling, and imagination-based planning

Methodology

This survey conducts a comprehensive literature review and analysis of world models and agentic AI for edge computing. It examines architectural foundations such as latent representation learning, dynamics modeling, and imagination-based planning, and illustrates proactive applications across multiple EGI scenarios. The methodology is conceptual and analytical, synthesizing existing work to provide a roadmap.

Key Results

The survey provides a conceptual foundation and practical roadmap for realizing intelligent, autonomous edge systems, highlighting how world models can enhance optimization under latency, energy, and privacy constraints.

Limitations

Safety guarantees remain an open challenge
Efficient training of world models is difficult
Constrained deployment on edge devices poses challenges

Resources

Paper