Large language models (LLMs) explained

What is a large language model?

A Large Language Model (LLM) is a type of AI trained on enormous amounts of text. The model learns to recognize patterns in language and can generate text, answer questions, summarize, translate, and reason.

How is an LLM trained?

1. Pre-training — predict the next word
2. Fine-tuning — adjust for specific applications
3. RLHF — Reinforcement Learning from Human Feedback

Known models

• Claude (Anthropic) — long context, strong reasoning, safety focus
• GPT-4o (OpenAI) — versatile, multimodal
• Gemini (Google) — integrated with Google services
• Llama 3 (Meta) — open-source
• Mistral — efficient European open-source alternative

Hallucinations

LLMs can produce factually incorrect information with great confidence. RAG (Retrieval-Augmented Generation) links an LLM to a knowledge base to reduce this.