English(EN) MemoryAthena: Adaptive Routing over Latent and Generated Memories

MemoryAthena系统探索为语言模型生成记忆

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-15 04:00

研究人员开发了MemoryAthena，一个新颖的系统，它探索在语言模型中使用除检索记忆之外的生成记忆。该系统采用三个路径：直接从显式记忆表（Engram）检索，从这些检索到的线索生成，以及直接从模型的内部状态生成。MemoryAthena学会选择性地干预生成记忆，从而提高问答和一般自然语言处理任务的性能。该系统（不包括冻结的骨干模型）包含约2.01亿个参数，并证明生成记忆可以有效地补充直接检索。 AI

影响引入了一种增强大型语言模型记忆的新方法，有望提高复杂推理任务的性能。

排序理由该集群包含一篇详细介绍新系统及其在基准测试上性能的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.CL 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; 该集群包含一篇详细介绍新系统及其在基准测试上性能的学术论文。[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.CL TIER_1 English(EN) · Mingyuan Li, Guangsheng Yu, Juyuan Zhang, Xu Wang, Zhibo Man, Haonan Zhang, Shaoxiong Ji · 2026-10-15 04:00

MemoryAthena: 跨越潜藏与生成记忆的自适应路由

arXiv:2609.25853v3 Announce Type: replace Abstract: Learned-memory methods store information in an explicit table and consume it through a separate reader, allowing addressing, storage, and reading to be modified independently. We study whether useful memory can also be generated…

本文如何被排名

报道来源 [1]

MemoryAthena: 跨越潜藏与生成记忆的自适应路由

相关实体

相关话题