Large Language Models (LLMs)

LLMs are Transformer-based deep learning models: they process petabytes of text data to generate, summarize, and translate human-like language using billions of parameters.

Large Language Models are the core engine for modern generative AI: these foundation models, like OpenAI's GPT-4 or Meta's LLaMA 2, leverage the Transformer architecture for scalable training. They utilize a self-attention mechanism to manage billions or even trillions of parameters, enabling nuanced understanding and generation of human language. This capability allows them to execute complex tasks: from writing code and summarizing technical documents to powering advanced conversational agents and providing real-time language translation. The technology is defined by its scale and its ability to generalize across diverse tasks with minimal task-specific fine-tuning.

https://en.wikipedia.org/wiki/Large_language_model