Work

Evaluating AI Models: Performance Validation Using Formal Multiple-Choice Questions in Neuropsychology