Intro
Embeddings, in the context of machine learning and natural language processing, are dense vector representations of discrete or categorical data (such as words) that capture semantic meaning.
They map items into a continuous, low-dimensional vector space of real numbers, where items with similar meanings are close to each other.
- Visualize embeddings: LLM Visualization
