English(EN) HiMA-Ecom: A Long-Horizon Memory Benchmark for Hierarchical Multi-Agent E-commerce Assistants

新基准HiMA-Ecom旨在解决多智能体AI中的长时记忆问题

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-28 04:00

研究人员推出HiMA-Ecom，这是一个旨在评估电子商务场景中分层多智能体系统的新基准。该基准包含超过22,000个实例，其中大部分侧重于智能体的记忆条目。该研究还提出了方差缩减组相对策略优化（VR-GRPO）来管理上下文预算并提高智能体训练效率。使用VR-GRPO和3B及7B模型的智能体在性能上优于DeepSeek-R1和DeepSeek-V3等现有模型，记忆机制缩短了收敛时间。 AI

影响为多智能体AI系统中的长时记忆引入了新的基准和优化技术，有望提高训练效率和性能。

排序理由该集群描述了一篇介绍用于AI智能体的新颖基准和优化方法的新学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 14 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该集群描述了一篇介绍用于AI智能体的新颖基准和优化方法的新学术论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, model release; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.AI TIER_1 English(EN) · Junxing Hu, Ai Han, Haolan Zhan, Pu Wei, Zhiqian Zhang, Yuhang Guo, Jiawei Lu, Zhen Chen, Haoran Li, Zicheng Zhang, Ling Li · 2026-10-28 04:00

HiMA-Ecom: 面向分层多智能体电子商务助手的长视界记忆基准

arXiv:2506.19846v3 Announce Type: replace Abstract: Long-horizon agentic systems become long-context systems by accumulation rather than through a single long document, since every step reassembles a prompt from role instructions, tool and sub-agent descriptions, recalled memorie…

本文如何被排名

报道来源 [1]

HiMA-Ecom: 面向分层多智能体电子商务助手的长视界记忆基准

相关实体

相关话题