When NumPy reshape shares memory and indexing does not
A reshaped feature array can alias the original data, while integer-array selection makes a copy. Inspect both before mutating a pipeline intermediate.
2026-10-22
ENGINEERING NOTES FOR THE CURIOUS · EST. 2026
Understand the
machine.
Open models. Real engineering decisions. Machine learning concepts that actually make sense. Practical references for people building with AI.
LIBRARY / 001
Start here.
Foundational explainers first. Primary sources, executable examples, and a clear account of what was tested.
When NumPy broadcasting gives you the wrong loss
A shape-compatible operation can still compare the wrong values. Diagnose a silent pairwise-error bug before it reaches your metrics.
2026-10-21
Split first, fit second: a preprocessing leakage check
Use a small scaler diagnostic to see whether held-out data is influencing training, then put the boundary inside a scikit-learn pipeline.
2026-10-21
BUILT TO GROW CAREFULLY
Four starting territories.
0 GUIDES
Open Models
Weights, licenses, and the decisions behind choosing a model.
1 GUIDEInference
Memory, latency, and the practical work of serving models.
3 GUIDESML Foundations
The ideas underneath the implementation.
0 GUIDESLLM Systems
Retrieval, evaluation, and software built around language models.