Placeholder chapter page. Sections are produced by the book-skills pipeline.
Sections
- 62.1 Why static deployment is obsolete
- 62.2 Catastrophic forgetting on the edge
- 62.3 Latent and quantized replay
- 62.4 Few-shot class-incremental learning (TinyOL)
- 62.5 Energy budgets for on-device training
- 62.6 Personalization on-device
- 62.7 Safety of self-updating models
Lab 62
implement latent-replay continual learning for a wearable model under a memory/energy budget.