A service-quality model is trained mainly on data from one location, although it will serve several locations. Human reviewers also use inconsistent labeling criteria. Which practices directly support a more responsible training dataset? Select 2 answers.

1. Document labeling criteria and audit inconsistent labels with qualified reviewers.
2. Duplicate examples from the dominant location until the dataset is larger.
3. Assess representation of the intended populations and operating conditions, and address meaningful coverage gaps.
4. Treat historical labels as unquestionably correct regardless of how they were produced.
5. Evaluate only the majority location because it contributes the most records.