arXiv · cs.AI · Nipun D. Pathirage, Kwan Srijomkwan · 9 小时前模型评分42
研究：GPT-5 mini 与 Grok 4 Fast 超九成外部知识问题会搜索，Mistral Small 仅两成
Confident Fools or Careful Searchers? A Behavioural Study of When LLMs Search Beyond Their Static Internal Memory
导读
一项新基准用 400 道选择题测试五个商用大语言模型（LLM，Large Language Model）何时该搜索、何时该直接作答，题目分为应直接回答的已知知识和训练截止后新闻类外部知识两类。
分类置信度
中 · 待人工复核
- 只有单一媒体来源，需要人工复核。
来源级别：媒体与个人
事件类型：研究发布
证据摘录
- “We introduce a benchmark of 400 multiple-choice questions that combines widely known inside-knowledge (IK) questions, which should be answered directly, with outside-knowledge (OK) questions about facts from news published after the models' training cut-offs (to our knowledge), which should trigger search or, without tools, abstention.”
来源：arXiv · cs.AI · arxiv.org