Large Language Model (LLM)

In Depth

Large language models are transformer-based neural networks with billions of parameters, trained on internet-scale text corpora. They learn statistical patterns of language and can generate coherent text, answer questions, translate languages, summarize documents, and write code. The 'large' refers to both the model size (parameters) and training data (terabytes of text). LLMs power tools like ChatGPT, Claude, and Google Gemini.

Examples

• GPT-4 with ~1.8 trillion parameters
• Claude 3.5 Sonnet by Anthropic
• Google Gemini Pro

Try It Yourself

Related Terms

Frequently Asked Questions