LLM Based Multi Agent

LLM-based multi-agent systems aim to enhance the capabilities of large language models by enabling collaboration among multiple agents, each specializing in different tasks or possessing unique knowledge. Current research focuses on improving agent communication, task decomposition strategies (like meta-task planning), and robust handling of diverse input modalities (including visual and textual data), often employing architectures inspired by internet protocols or assembly-line paradigms. This field is significant because it addresses limitations of single-agent LLMs in complex tasks, paving the way for more sophisticated applications in areas such as data science automation, software development, and urban planning.