Mistake Clinic

Module-Specific Mistake Radar

Mistake	Signal	Repair
Shape reasoning by trial	Silent broadcasts	Write every tensor shape
One lucky seed	Large run variance	Repeat and report dispersion
Mask leakage	Future tokens influence output	Assert the mask with a toy sequence
Fine-tune by default	No frozen baseline	Compare prompting and retrieval
Quality-only compression	Runtime regressions hidden	Measure target hardware

Practice Mistake Checks

Check shapes, masks, splits, seeds, ablations, target-device latency, and capability boundaries before review.

Evidence and review

Repair one failure and rerun all affected evidence. 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.