Capstone Thread

Add an architecture note, tokenizer analysis, training/evaluation run, adaptation comparison, ablation table, inference profile, and capability-boundary statement to the capstone.

Evidence and review

A reviewer must reproduce one run and challenge one architectural claim. Claims must trace to versioned code and artifacts. Screenshots, unexplained generated prose, and a single favorable run do not pass.

Oral defense

Explain the mechanism, identify the most dangerous assumption, interpret one failure from raw evidence, and name the condition that would reverse your decision.

Source backbone

Use Dive into Deep Learning, PyTorch documentation, primary papers, and Designing Data-Intensive Applications selectively.