Reproduce the baseline¶
No variant counts until the reference is reproducible.
Working notes¶
- Freeze the inputs, split, seed policy and evaluation procedure.
- Run the reference before changing the method.
- Include a negative control to expose an evaluator that rewards everything.
- Keep the baseline artifact beside each candidate result.
Source: course evidence synthesis, §3–4. These are selected observations, not general performance guarantees.
Before the course¶
- [TODO: select a compact worked example and its verification artifact.]
- [TODO: rehearse the live walkthrough and prepare a fallback.]