Add benchmark data (23/37): evaluation/llm/annotation_schema.json

Add benchmark data (22/37): evaluation/llm/question_bank_v3.json

Add benchmark data (21/37): evaluation/llm/question_bank.json