what are agents and multi agents in Gen AI?

Описание: Please watch: Evolution of RAG Systems: RAG → GraphRAG → KAG + CAG (Caching) :   • Evolution of RAG Systems: RAG → GraphRAG →...  

what are agents and multi agents in Gen AI?
===========================================================
provides a comprehensive overview of how Large Language Models (LLMs) evolve into more sophisticated systems like agents and multi-agents by integrating various augmentation techniques. It explains the progression from basic LLM functionality to advanced concepts such as Retrieval Augmented Generation (RAG) Graph RAG and Knowledge Augmented Generation (KAG), leading up to the core idea of intelligent, tool-using agents and their collaborative multi-agent counterparts.


Here are the key takeaways:
========================

Evolution of LLMs (0:06-2:31): The video traces the development from basic LLMs to enhanced versions like RAG, Graph RAG, KAG, and CAG, each adding layers of external context, hierarchical understanding, knowledge graphs, or caching to improve output.
Agents: LLMs with Decision-Making (2:37-5:44): An agent is an LLM endowed with the ability to think, perceive its environment, decide, and act. This involves understanding available tools, selecting the most appropriate one for a given query, and then executing actions to provide a relevant response.
Multi-Agents: Collaborative Intelligence (5:49-8:08): Multi-agents represent a system where different specialized agents (e.g., research, web, compute, medical) coordinate and communicate to solve complex problems. A supervisor agent often routes queries to the most suitable agent, allowing for a collaborative and comprehensive approach, as demonstrated by the "hub-to-spoke" framework.

Hashtags:
#RAGSystems #GraphRAG #KAG #CAG #LLMs #GenAI #AIArchitecture #KnowledgeGraphs #RAG #agents #multiagent #multiagentsystems #agenticai #genai #generativeai