arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2603.14864cs.CL
收藏

购物助手：面向长期偏好引导的电子商务任务的LLM智能体基准测试与训练

Shopping Companion: Benchmarking and Training LLM Agents for Long-Horizon Preference-Grounded E-Commerce Tasks

Alibaba International Digital Commercial Group（阿里巴巴国际数字商业集团）

Zijian Yu, Kejun Xiao, Huaipeng Zhao, Tao Luo, Xiaoyi Zeng

阅读 PDF
arXiv 页面 ↗HTML 原文 ↗

AI总结：

针对电子商务中缺乏长期偏好感知购物任务基准和细粒度训练监督的问题，提出了Shopping Companion Bench基准和免标注工具级奖励方法，有效提升了LLM智能体的偏好捕获与任务性能。

摘要：

在电子商务中，LLM智能体在推荐、预算管理和捆绑销售等购物任务中展现出潜力，其中从长期对话中准确捕捉用户偏好至关重要。然而，进展受到两个关键挑战的限制：（1）缺乏评估长期偏好感知购物任务的基准，（2）缺乏用于购物智能体训练的细粒度监督。为了填补基准空白，我们引入了Shopping Companion Bench，这是一个新颖的基准，包含两个需要跨会话偏好记忆的购物任务，基于超过120万真实商品的产品池。我们的分析进一步指出了该基准上失败的两个主要来源：偏好幻觉导致的级联错误，以及未能充分验证产品属性是否符合用户需求。为了解决这些失败模式，我们设计了免标注的、工具级奖励，为每次工具调用提供过程监督，从而缓解了长期任务中的奖励稀疏问题。实验结果表明，即使是GPT-5等最先进模型，成功率也低于70%，凸显了我们基准的难度。值得注意的是，我们微调的轻量级4B模型在偏好捕获和任务性能上均持续优于强基线，表明我们奖励设计的有效性。

英文摘要：

In e-commerce, LLM agents show promise for shopping tasks such as recommendations, budget management, and bundle deals, where accurately capturing user preferences from long-horizon conversations is critical. However, progress is limited by two key challenges: (1) the absence of benchmarks for evaluating long-term preference-aware shopping tasks, and (2) the lack of fine-grained supervision for shopping agent training. To fill the benchmark gap, we introduce Shopping Companion Bench, a novel benchmark comprising two shopping tasks that require cross-session preference memory, grounded in a product pool of over 1.2 million real-world items. Our analysis further identifies two major sources of failure on this benchmark: cascading errors caused by preference hallucination, and insufficient verification of product attributes against user requirements. To address these failure modes, we design annotation-free, tool-wise rewards that provide process supervision for each tool call, alleviating reward sparsity in long-horizon tasks. Experimental results demonstrate that even state-of-the-art models such as GPT-5 achieve success rates below 70%, highlighting the difficulty of our benchmark. Notably, our fine-tuned lightweight 4B model consistently outperforms strong baselines in both preference capture and task performance, suggesting the effectiveness of our reward design.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

↑