opencode_meta_cognitive_dataset.stats.json

{
  "total_examples": 500,
  "positive_examples": 251,
  "negative_examples": 249,
  "framework_distribution": {
    "diagnostic_hierarchy": 125,
    "evidence_based": 100,
    "scope_boundary": 75,
    "conciseness": 100,
    "question_before_answer": 100
  },
  "avg_instruction_length": 22.796,
  "avg_output_length": 104.712
}