LLM-Based Autonomous Agents: A Systematic Review and Critical Synthesis of Architectural Paradigms, Memory Models, Planning Strategies, And Operational Limitations
Large Language Models (LLMs) have enabled the emergence of autonomous AI agents capable of reasoning, planning, tool use, and iterative decision-making. Despite rapid development, the field remains architecturally fragmented, with limited conceptual clarity regarding memory integration, planning mechanisms, and oper...