2 papers
cs.CE2026
Historical Backtesting for Scientific Question Discovery: A Protocol and Astronomy Pilot
Hui Mao
Systems that generate scientific research questions are evaluated today by expert scores, LLM-as-judge ratings, or curated case studies -- all subjective, none falsifiable. We form…
cs.DL2026
Evidence-Based Scientific Question Discovery: A Framework with Historical Backtesting
Hui Mao
Current AI systems are optimized for answering questions; the scientific enterprise is bottlenecked earlier, at discovering the questions worth investigating. We present a framewor…