Skip to content
Learning, then using a model
Separate learned parameters from information supplied with a request.
How it works
Learning
Training adjusts parameters to reduce a loss or increase a reward. The objective, data and optimisation procedure shape what is learned. At inference, the model uses its parameters and available context to compute an output.
Check my understanding
Before answering, form your own explanation.
Adding a sentence to the instructions necessarily retrains the model.
Takeaway
Changing instructions changes the input; it does not necessarily retrain parameters. Supervised learning, self-supervised learning and reinforcement learning describe methods, not architectures.
Go deeperFrom concept to practicePut it into practice
Scientific sources
- Vaswani et al. · 2017Attention Is All You NeedarXiv v7 · 2023Original publication
- Hafner et al. · 2023Mastering Diverse Domains through World ModelsDreamerV3 · arXiv v2 · 2024Original publication
Teaching synthesis of the cited sources. Published results remain tied to their tasks, protocols and budgets.