arXiv · cs.AI · Bharath Sivaram Narasimhan, Karthik R. Narasimhan · 4 小时前模型评分46
τ-Rec-Books：面向智能体推荐系统需求引导的披露控制基准
$\tau$-Rec-Books: A Disclosure-Controlled Benchmark for Requirement Elicitation in Agentic Recommender Systems
导读
研究者发布 τ-Rec-Books 基准，固定用户请求、只改变需求披露时机（一次性给出、被问才说、部分永不透露），用 2 万本书目生成 450 个任务评测 人工智能（AI，Artificial Intelligence） 智能体的需求引导能力。
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
- “To evaluate the ability of AI agents at this elicitation process, we introduce $\tau$-Rec-Books, a benchmark that holds each user request fixed while varying only when the user reveals their requirements: all at the start (volunteer), one at the start and the rest when asked (mixed), or some never (hidden).”
- “We generate 450 tasks over a catalog of 20,000 books from Wikidata, develop a rule-based simulated user to answer the agent, and score recommendations against catalog metadata without an LLM judge.”
- “We release the catalog pipeline, task generator, simulated user, validators, reference agents, and all agent traces at this https URL .”
来源：arXiv · cs.AI · arxiv.org