LLM, Model, Token, Context Window
This article explains the fundamental concepts of Large Language Models (LLMs), distinguishing between the general technology (LLM) and specific instances (models) like GPT-4o or Claude Sonet. It details how text is broken down into tokens, which are the basic units LLMs process, and clarifies that one word does not always equal one token. The piece also defines the context window as the temporary memory space an LLM uses during inference, encompassing prompts, chat history, and attached files, which is crucial for maintaining conversational memory and processing larger amounts of data. AI
IMPACT Clarifies core LLM concepts, aiding developers in understanding model capabilities and limitations.