Large Language Model (LLM)

Quick Definition: Large Language Model (LLM)

A Large Language Model (LLM) is a form of generative AI model trained on massive corpora of text using self‑supervised learning. It is specialized in tasks involving natural language, such as text generation, summarization, translation, and conversational response.

LLMs are typically based on transformer architectures and consist of billions to trillions of parameters.

Key points:

• They use self‑supervised learning: the model predicts masked or next tokens within large text datasets.
• They can be fine‑tuned or adapted for specific domains or tasks.
• LLMs may produce coherent but sometimes factually incorrect outputs (“hallucinations”).