Take the 01 - 06 pages.
     what needs to be expanded on these ...addition pages for more depth, description and teaching.
    
    
    add target to inner page links
     
     
For example


The Training Pipeline
1 Raw data collection — web crawl, books, code, papers
2 Data cleaning — filter, deduplicate, normalize, remove toxic content
3 Tokenization — text converted to token IDs the model can process
4 Pre-training — billions of gradient descent steps across the full dataset
5 Fine-tuning — curated examples shape specific behaviors
6 RLHF — human feedback trains the reward model; model learns to maximize it
7 Evaluation & safety testing — red-teaming, benchmarks, capability assessment
8 Deployment — weights frozen, model moves to inference infrastructure

     
     
ALSO once complete review  NAV and bread crumbs for new AI-Stak pages