精选理由
Perplexity AI 搞了个新评估方式 WANDR，不用固定答案打分，而是实时对证据，适合处理时效性事实。
WANDR不再使用预定义的黄金解决方案来评分，而是动态重新抓取每个引用的页面。它逐条核对每个声明与引证证据的一致性。这种方法适用于需要处理随时间变化事实的任务。
原文 · Perplexity
Instead of grading against a gold solution, WANDR re-fetches every cited page and checks each claim ...
Instead of grading against a gold solution, WANDR re-fetches every cited page and checks each claim against the underlying evidence. This allows certain tasks to contain time-varying facts. 💬 3 🔄 1 ❤️ 8 👀 1597 📊 4 ⚡