Original Summary

Open AI research experiment and LLM benchmark testing whether curiosity-driven warm-ups improve Claude reasoning accuracy. Preregistered, reproducible AI evaluation across Claude, Claude Code, and API sessions. Join the 10-minute volunteer study. No coding required.

  • 情报分类：商业与市场研究
  • 分类依据：内容涉及商业、投资或市场动态
  • 信息来源：GitHub · AI 新项目
  • 发布时间：2026/10/3 08:04:41