Glossary

A diffusion model creates images by refining random noise step by step into realistic results, forming the backbone of modern generative AI tools.

A context window defines how much text an AI model can process at once, shaping memory, reasoning depth, and performance in large language models.

Transfer learning allows AI models to reuse knowledge from previous tasks, saving time, improving performance, and reducing training costs.

Pruning removes unnecessary neurons or weights from neural networks, making AI models smaller, faster, and more efficient without losing accuracy.