◆ Definition
Training
The process of building a model: showing it enormous amounts of data and adjusting its parameters until its predictions improve. Followed by post-training (instruction tuning, preference optimization) that shapes raw capability into useful behavior.
Explained for kids
How AI learns — showing it millions of examples and letting it correct its own mistakes over and over until it gets good at spotting the pattern.
Learn more in these free lessons
- Pre-training & Fine-tuning Pipeline · AI Deep Dive
- Scaling Laws · AI Deep Dive
- The Analyst Role Shifts · AI for Cybersecurity
- The Compute Stack · AI Deep Dive
- The Economics of Intelligence · AI Deep Dive
- Hallucination: Mechanisms & Mitigations · AI Deep Dive