SECAI Core · phase 2 of 15
Pre-trained model
An AI model that has already been trained on a large, general dataset and can be reused or fine-tuned for specific tasks
The Explain card
- Plain English
- A pre-trained model has already been trained on a large, general dataset. You download it and reuse it as-is or fine-tune it for your own task instead of training from scratch.
- Example
- A team downloads an open-source image classifier from a public model hub and fine-tunes it to spot tampered identity documents.
- Why it matters
- You inherit everything baked into that model: its biases, memorised data and any backdoor planted upstream. Pre-trained models are a supply chain, and they deserve the same scrutiny as third-party code.
- Hook
- Someone else's homework, with someone else's mistakes inside.
Where it sits in the deck
Phase 2: How Models Are Built: Architectures and Learning Mechanics
With the paradigms named, zoom in on the architectural building blocks that turn data into trained artifacts.
