1
AI不再用完即忘：华为诺亚开源MindMemOS，记忆和Skill一起进化
华为诺亚开源记忆系统，让AI记住你喜欢什么、现在用啥，还带着技能一起升级。
面向AI Agent的可迁移、自演进记忆操作层MindMemOS
2
Memory Layer: Train the In-Model Cache for Recommendation Models
将缓存训练融入推荐模型内部，这篇论文提出的Memory Layer有望大幅提升推荐系统效率与性能。
arXiv:2607.25110v1 Announce Type: cross Abstract: Early ranking stages in recommendation systems precompute item embeddings and cache them in-model fo…
3
MemoHarness: Agent Harnesses That Learn from Experience
智能体如何从经验中持续进化？这篇论文提出MemoHarness框架，让Agent学会自我迭代。
arXiv:2607.14159v1 Announce Type: cross Abstract: An agent harness is the external control layer that turns a base LLM into an executable agent by man…
4
AI's memory. On your machine, under your control
本地控制AI记忆的开源项目，每次写入需你批准，杜绝无声隐私窃取。
Article URL: https://github.com/EXXETA/exxperts Comments URL: https://news.ycombinator.com/item?id=48884002 Points: 1 # Comments: 1
5
Show HN: Memory Magico – CLI based memory, wiki and deterministic sprints
厌倦GitHub issues和Mira？这款CLI工具将memory、wiki与确定性sprint管理合体，让你高效追踪长sprint和bug排查，并让agent负责到底。
hey, I got tired of using GitHub issues and Mira to manage sprints and issue tracking so this is a tool I have been using for a while in a few of my p…
6
Show HN: Memoriq – Private AI Memory for ChatGPT, Claude, Gemini and Grok
为ChatGPT、Claude等AI大模型提供私有记忆，保护隐私的同时提升对话连贯性。
Article URL: https://memoriq.me/ Comments URL: https://news.ycombinator.com/item?id=48504064 Points: 2 # Comments: 0
7
Agent Memory SDK – persistent memory for AI agents
为AI代理赋予持久化记忆的开源SDK，让智能体不再“失忆”，值得开发者关注。
Article URL: https://github.com/gharibyan/agent-memory Comments URL: https://news.ycombinator.com/item?id=48494184 Points: 2 # Comments: 0
8
Show HN: Local Context and Memory Stack
本地AI记忆栈开源项目，帮你管理上下文与持久记忆
Article URL: https://github.com/supermemoryai/supermemory Comments URL: https://news.ycombinator.com/item?id=48479048 Points: 1 # Comments: 0
9
Efficient and accurate neural-field reconstruction using resistive memory
突破性硬件AI：用电阻式存储器实现高效精确的神经场重建，Nature发布。
Nature, Published online: 10 June 2026; doi:10.1038/s41586-026-10646-w A co-optimized AI hardware–software system using resistive-memory computing imp…
10
How I Gave My AI Agent a Real Memory Without Touching Its Internals
不触碰AI内部结构，用Memory Sidecar三层架构给智能体添上持久记忆，版本3.1.1稳定可靠。
As someone who runs multiple AI agents for day-to-day tasks, I kept hitting the same wall: every new conversation was a total memory wipe. I'd ask my …
11
Show HN: Standing Questions – agent memory that stores questions, not answers
一种颠覆性的 agent 记忆设计：只存问题不存答案，让信息随版本自然更新
Article URL: https://github.com/Rocco-alt/standing-questions Comments URL: https://news.ycombinator.com/item?id=48460257 Points: 1 # Comments: 0
12
The Cheapest Way to Self-Host Memos in 2026
每月3美元一键部署，30秒获取HTTPS实例，2026年自托管Memos的最省钱方案。
Last updated: June 2026 Memos is the lightweight, open-source notes app a lot of people land on after getting tired of flomo, Google Keep, or a paid N…
13
Dual-Cluster Memory Agent: Resolving Multi-Paradigm Ambiguity in Optimization Problem Solving
LLM解优化问题常陷多范式歧义，DCM-Agent双簇记忆机制精准破解，值得技术研究者关注。
arXiv:2604.20183v2 Announce Type: replace Abstract: Large Language Models (LLMs) often struggle with structural ambiguity in optimization problems, wh…
14
MemORAI: Memory Organization and Retrieval via Adaptive Graph Intelligence for LLM Conversational Agents
最新论文提出MemORAI方法，用自适应图智能优化LLM对话代理的记忆组织与检索，提升长期对话能力。
arXiv:2605.01386v2 Announce Type: replace Abstract: Large Language Models (LLMs) lack persistent memory for long-term personalized conversations. Exis…
15
eMoT: evolving Memory-of-Thought via Symbolic Anchoring and Memory Corrosion
一种新记忆进化机制eMoT，通过符号锚定与记忆腐蚀，在数学推理任务中实现17.6%准确率提升，Game of 24达100%。
arXiv:2606.02054v1 Announce Type: new Abstract: While Large Language Models (LLMs) achieve impressive performance on multi-step reasoning tasks, their…
16
Hermes Agent 加入工具搜索，“AI 养马”更省 tokens
腾讯开源Agent Memory，联合Hermes Agent工具搜索，让token消耗暴降61%，准确率从49%飙升至74%
IT之家 5 月 30 日消息，科技媒体 marktechpost 昨日（5 月 29 日）发布博文，报道称 Nous Research 为开源 Hermes Agent（Hermes 智能体）加入 Tool Search（工具搜索）， 通过按需加载缓解 MCP（模型上下文协议）工具占满上下文的问题…
17
MIT's MeMo lets teams swap in a better LLM without retraining — and performance jumps 26%
不用重新训练就能替换大语言模型，MIT的MeMo方案让性能飙升26%，打破向量数据库局限
Enabling LLMs to acquire new knowledge after training remains a major hurdle for enterprise AI — current solutions are either too expensive, too slow,…
18
Memory Architectures for Multi-Turn Text-to-SQL: A Benchmark and Empirical Study
多轮对话中记忆架构如何影响Text-to-SQL性能？基准与实证研究揭示关键设计差异。
arXiv:2605.26394v1 Announce Type: new Abstract: Multi-turn Text-to-SQL is central to enterprise analytics yet remains predominantly evaluated in singl…
19
Memorial Day Tech Deals: Sony, Apple, Beats (2026)
索尼、苹果、Beats耳机等热门科技产品促销，价格接近历史最低，不容错过。
Lots of our most-recommended headphones, power banks, and other gadgets are on sale for Memorial Day.
20
Evo-Memory: Benchmarking LLM Agent Test-time Learning with Self-Evolving Memory
提出Evo-Memory基准，全面评估LLM Agent通过自演化记忆进行测试时学习的能力。
arXiv:2511.20857v2 Announce Type: replace Abstract: Statefulness is essential for large language model (LLM) agents to perform long-term planning and …