Protocol
Evaluation, Ablations and Responsible Claims
Audit a model-comparison claim that appears impressive. Rebuild the evaluation so that a null or negative finding would be a valid and publishable outcome.
35–60 active hours60–90 min sessionsIntermediate
Mastery contract
Core protocol · Standard 1.0 · Audit a model-comparison claim that appears impressive. Rebuild the evaluation so that a null or negative finding would be a valid and publishable outcome.
Loading your learning records…
Reasoning — not yet passed
The hypothesis and evaluation protocol are written before outcomes, including what would disconfirm the claim.
Reliable implementation — not yet passed
Evaluation checks prediction-label alignment and correct resampling units with explicit edge-case fixtures.
Reproducible experiment — not yet passed
Report effect sizes, uncertainty, all ablations and subgroup sample counts under comparable budgets.
Defense and handoff — not yet passed
Write a bounded conclusion that identifies the strongest baseline, weakest evidence and one untested deployment assumption.
All four criteria must pass. Activity completion does not satisfy them.
An independent reviewer reruns your work and varies something you did not rehearse: a fresh input, a different working directory, or a declared edge case.
Pilot decisions are provisional, not external credentials. For an appeal or sensitive artifact, contact your designated pilot operator with the submission ID. Export your assessment history.
What you will do
6 total
Define one falsifiable hypothesis, primary metric, resampling unit and disconfirming result; explain why 3 training seeds do not quantify dataset uncertainty.
Implement an evaluation harness with frozen predictions, per-example IDs, subgroup slices and paired bootstrap differences.
Compare a constant baseline, a simple model and a stronger model; run one ablation for each claimed improvement and retain every result.
Debug a metric with reversed positive labels and a subgroup comparison with too few independent examples.
Write a 500–900 word technical report linking the reasoning, code, measurements, and limitations; include the project decision below.
Assess all 4 mastery criteria against saved artifacts, request an independent review, and repeat each failed criterion on a fresh example.
Essentials
A reading session alone never satisfies a mastery criterion.
Store source references, assumptions, and artifact paths beside every result.
Keep an untouched check case that differs from the worked example.
Record all attempts, including failures and results that contradict your prediction.
Use the stated comparison conditions; document every deviation before drawing a conclusion.
Protocol guardrails
Do
- +State the expected result before running the comparison.
- +Keep one minimal reproducible failing case when debugging.
- +Record environment versions and the exact command used.
- +Compare explanations with saved intermediate values.
- +Ask a reviewer to challenge the weakest assumption.
Don't
- ×Do not copy a worked solution and present it as an independent implementation.
- ×Do not tune against held-out evaluation outcomes.
- ×Do not report only the best seed or discard inconvenient runs.
- ×Do not equate elapsed hours or a completed run with a passed assessment.
- ×Do not conceal reduced-scale experiments behind claims about the original full-scale result.
Protocol authorship
Written by NuthinButta as instructional design. The teaching sources are the official references linked in each milestone; the exercises, workload and pass thresholds are ours, not their authors'.