When computers see
Instrument the 3x3 plus-sign classifier
Page 6 adds signals that distinguish bad input from component failure in the 3×3 plus-sign image classifier.
1Learn the idea
Read
Emit stage signals
Instrument the 3×3 plus-sign image classifier so a run records enough structure to debug offline: counts, latency if relevant, pass/fail of accuracy on a fixed four-image set, and a stable stage name. Redact secrets and raw credentials from every event.
Read
Emit and assert
truth=['plus','plus','other','other']; pred=['plus','other','other','other']
tp=sum(a==b=='plus' for a,b in zip(truth,pred)); fn=sum(a=='plus' and b!='plus' for a,b in zip(truth,pred))
print({'tp':tp,'fn':fn,'recall':tp/(tp+fn)})
Expected evidence: stage evidence for observability. Prefer JSON or structured text you can grep in CI over prose logs for computer-sees.
Read
Lock signals with a regression test
Turn one historical failure—especially overfitting to one plus orientation—into a test that fails if the signal disappears for the 3x3 plus-sign classifier. Observability without a failing test is optional decoration; observability with a test is part of the computer-sees artifact.
Read
Lab notebook: signal schema
Draft a three-field event for the 3x3 plus-sign classifier: stage, ok, and one domain field derived from accuracy on a fixed four-image set; feature activations printed. Add fixture_id or docs_version when content can change. Explicitly list fields that must never appear (tokens, passwords, raw prompts) because claiming 'computer vision' production readiness from 3×3 toys is in scope for this lab.
Wire one assertion that fails if the 3x3 plus-sign classifier event is missing after a run. Observability that cannot fail a test will not survive contact with a busy computer-sees repository.
Read
Worked judgment
Imagine a teammate opens only your event stream after a bad deploy. Could they tell whether plus, blank, and distractor 3×3 grids was wrong, whether overfitting to one plus orientation, or leaking test grids into template design returned, or whether claiming 'computer vision' production readiness from 3×3 toys slipped through? If not, rename fields until those three stories are distinguishable.
Read
Why this stage matters for the 3x3 plus-sign classifier
At the testing and observability stage for computer-sees, the job is narrower than finishing a product demo. You are creating one progressive evidence piece about plus, blank, and distractor 3×3 grids that later pages inherit without redefining success. Keep that fixture small enough to inspect by hand, keep outputs copy-pasteable as text, and refuse to narrate this baseline as if it were a production SLA: pixel-sum threshold accuracy on the same four images.
For this page specifically, success looks like a structured event schema locked by a test while still centering the user decision to classify tiny binary images as plus vs not-plus with inspectable features. If you cannot point to a file, command, or assertion that proves that for the 3x3 plus-sign classifier, stay on this page instead of advancing.
Glossary: computer vision · Glossary: convolutional neural network
Go deeper
Before you start
Why this matters
Write the single log line or metric event that would tell you whether a bad result came from input vs implementation for the 3x3 plus-sign classifier. If your line could not tell them apart, redesign it before coding.
In the wild
See how this idea shows up as a product and a company — then come back to the lesson. Skills transfer across vendors.
Related lessons
Check your understanding
Page assessment
Answer from memory. Completion is saved from this evidence, not from opening the next page.
All responses are required.