Research Papers

Multi-Model Cross-Validation for Zi Wei Dou Shu AI Interpretation:
A Reproducible Benchmark and Transparency Framework

We present the first publicly documented AI benchmark for Zi Wei Dou Shu （紫微斗數） interpretation. Our multi-model cross-validation architecture uses OpenAI / Anthropic / Google Gemini governance benchmark references to independently validate analyses, achieving +6.2 points over single-model baselines on our six-dimensional rubric. We release starnum-bench: 37 sanitized charts with scoring rubric, evaluation script, and April 2026 baseline results.

Categories: cs.AI · cs.CY (Computers and Society)

@techreport{starnum2026preprint,
  title     = {Multi-Model Cross-Validation for Zi Wei Dou Shu AI Interpretation:
               A Reproducible Benchmark and Transparency Framework},
  author    = {{Starnum Research Team}},
  year      = {2026},
  month     = {April},
  institution = {starnum.com.tw},
  type      = {Preprint},
  url       = {https://starnum.com.tw/papers/},
  note      = {Version 1.0. Preprint available at starnum.com.tw/papers/}
}