Update README: restore config/task tables

Update README: remove scielo-bench, simplify text, update citation and journal count

Ship SapBERT-linked exam CUIs (predicted_cuis + per-category) so exam-recommendation F1 is reproducible

Ship SapBERT-linked DDx CUIs (predicted_cuis + per-item cuis) so differential-diagnosis F1 is reproducible

Add 3-judge diagnosis verdicts (gpt-oss+qwen+llama) + consensus to results_diagnosis

Exam recommendation: report macro-F1 (match paper); rename exam_micro_f1 -> exam_macro_f1

Exam recommendation: report macro-F1 (match paper); rename exam_micro_f1 -> exam_macro_f1

Add ClinicalBr v1: corpus + 4 task benchmarks + reference results (PT/EN)

initial commit