rag · 知识库活文档

更新：R129：0 件主分类净增 + 8 件增量 + 1 件新 arXiv ID（分类错位）；新增成本—质量交换轴与委托链权限断裂轴。

1. 现状全景

R129（接续 R128-R82 累计 48 轮）：宿主机增量清单驱动，只处理清单 8 份材料，未做全库扫描；2 次 web_search 补最新进展。本轮完整论述见 archive/rag-changelog.md（R129 段）。

增量 8 件：A 组 5 repo_cards（takeshy/obsidian-local-llm-hub + api-evangelist/pienso + Ablation-Tool/ablation + devrochy/agent-core + crmne/ruby_llm）｜B 组 Jay 技术博客动态（2026-10-26 13:35，6 条：2 已核 200 / 4 搜索线索待核）｜C 组 Jay CSDN 线索（2026-10-25 16:20，2 条：1 已核 200 / 1 条 521 未重试）｜D 组 1 paper_card（2001.08361 Scaling Laws）。
新增 arXiv ID 1 件：2001.08361（S2 9390，OpenAlex 1747，主分类 llm-infra）。RAG 主分类净增 = 0——继 R128 的 BLIP-2 / Skip-Thought / BigBird 后连续第二轮 主题: rag 标签命中但主分类非 RAG 的错位件，登记为分类噪声，不进 SOTA 主轴。新增 arXiv DOI 1 件（10.48550/arxiv.2001.08361）+新增非 arXiv 学术 DOI 1 件（10.18653/v1/2026.findings-acl.648）。新增 CVE 0 件。
最强单件证据：Reflective RAG（ACL Findings 2026, 2026.findings-acl.648, Wu et al., pp.13252–13267，web_search 补入）——"自评估驱动的策略优化"：SFT 基于内部不确定性自校正标签学反思信号；RL 用反思优化策略 + 动态 KL 正则稳定训练；自述在 5 个知识密集 QA 基准上一致优于强 agentic 基线，多跳推理泛化更好。这是本库第一次把"自评估信号"做成可训练目标的 agentic RAG 工作。
工业侧交叉证据：Taskade 引 Infragistics 2026-07 的 100 查询对照（RAG vs MCP 式 agentic retrieval：成本 $0.0047→$0.1691 +3500%；总延迟 12,715ms→31,932ms +151%；首 token 6,314ms→29,372ms +365%；答案胜率 38%→52% +14 点；引文数 2.2→4.5 +2.3）；OvalEdge 引 CSA 2026 称逾半数采纳 NIST AI RMF 的企业仍报多 agent 治理缺口（"agentic sprawl"），失效模式为"下游 agent 用了自己的权限而非用户的权限"。
R129 六轴：① 成本—质量交换轴 ② 委托链权限断裂轴（2610.30418 复合结论级越权落到"agent 间委托"机制）③ 自评估可训练轴（Reflective RAG）④ 分类错位连续轴 ⑤ 多租户权限基线轴（devrochy/agent-core per-tenant AgentSpec + Postgres RLS）⑥ 局部 RAG 记忆策略轴（短期记忆三策略 + PostgresSaver checkpointer）。
fresh 来源：organized/repo_cards/{4652,4649,4638,4636,4380}.md｜inbox/jay/2026-10-26T1335-blogs-discovery.md｜inbox/jay/2026-10-25T1620-csdn-discovery.md｜organized/paper_cards/1128-2001-08361.md｜web_search 补 2 件（Reflective RAG 2026.findings-acl.648；Taskade 成本对照 / OvalEdge 治理缺口）。

2. 关键工作与脉络（R129 增量 8 件 + 四项判断）

A 组 5 repo_cards（均首次入本文档视野） - takeshy/obsidian-local-llm-hub（90⭐ TS MIT，agent/app，2026-10-26 experimental）——本轮唯一真正可用的 RAG 系统：Obsidian 本地 hub（LLM 对话 + vault 工具 + MCP servers + RAG + 工作流自动化 + 加密 + 编辑历史，全私有无云）。 - api-evangelist/pienso（0⭐ engineering，framework，待 LLM 分类，2026-10-26）。 - Ablation-Tool/ablation（223⭐ Python NOASSERTION，engineering，framework，AIARE 框架，2026-10-25，待 LLM 分类）。 - devrochy/agent-core（0⭐ Python MIT，agent，framework，2026-10-25）——per-tenant AgentSpec + Postgres RLS + MCP 式插件工具 + 三层记忆 + 幂等 + guardrails + OpenTelemetry，本轮唯一触及多租户 RAG 权限模型者；0⭐、experimental、无提交历史。 - crmne/ruby_llm（4483⭐ +11 周增，Ruby MIT，multimodal，research，2026-10-25）——成熟度最高的语言绑定件，非 RAG。

B 组 Jay 技术博客与开源动态（6 条） - 已核 200｜IBM《agentic RAG pipeline with multi-agent AI workflows》：CrewAI 四智能体骨架（answer_generator / product_answer_generator / reqs_retriever / products_retriever）+ 意图分类层（产品型 / 需求型 / 定价型 / 无关型路由）+ 混合数据层（结构化规格走 Presto on watsonx.data 的 SQL 表，字段 input_voltage_min/max、output_voltage、power_rating、efficiency、isolation_ratio；语义检索走向量库）。 - 已核 200｜swyx《Claude Mods》：Claude Code 内 TS/JS 事件驱动中间件，可改渲染与执行流；Anthropic 内置六个 mod（agents-md、diff、plugin-authoring、sec-default、telemetry、you-should-know，默认关闭）。 - 搜索线索待核｜NirDiamant/RAG_Techniques、PacktPublishing/Building-Production-Ready-AI-Agents（Chapter-01~19 含 Responsible-AI / Tool-Calling / RAG / Advanced-RAG / ReAct）、microsoft/ai-agent-runbooks（企业 RAG：SharePoint/OneDrive/Dataverse/Azure SQL/Fabric 接入 + 文档级权限 + 引用链）、Java Revisited《30 Must-Know AI Engineering Concepts for 2026》。 - ⚠️ B 组无一件进入量化结论：IBM 的 95% 准确率是单案例（工业电源目录）自述，无评测集、无基线对照、无复现路径；swyx 的 "You Should Know" 是作者个人工作流自述，且"审批策略从 base 分支读取"本次未在取回正文中定位到对应段落；后四条仅目录/摘要印象。价值在架构范式与交互范式。

C 组 Jay CSDN 线索（2 条） - 已核 200｜《拥有短期记忆的 RAG Agent》：短期记忆三策略与图结构——条件分支汇集节点 passthrough、消息条数阈值 count_messages 判定、summarize_conversation 中 trim_messages(strategy="last", start_on="human", end_on=("human","ai")) 裁剪 + RemoveMessage 删除、工具消息（msg.type != 'tool'）排除，基于 StateGraph + PostgresSaver checkpointer 装配（节点、边、条件边、compile）。原始出处 https://machinelearningwithme.com/agent/chapter05/short-term-memory-rag-agent/ ，AtomGit 转载镜像，前序章节不可见；LangGraph/PostgresSaver 版本号、依赖安装与 Postgres 配置均未确认，代码未复现。 - 未核｜《多模态智能体与 RAG 落地》521，按要求不重试；抓取内容夹杂无关侧栏文本、疑似拼接/模板化痕迹，暂不推荐。

D 组 1 paper_card（分类错位）｜2001.08361 Scaling Laws（S2 9390，OpenAlex 1747，影响力被引 645，OA green，主分类 llm-infra，主题标签 rag）。TLDR："更大的模型显著更具样本效率，最优算力高效训练是在相对适中的数据量上训练非常大的模型，并在远未收敛时显著提前停止。"讲预训练规模律，与检索增强生成无实质关系，不计净增、不进 SOTA 主轴。

判断一 · agentic 检索成本—质量交换第一次有了对照表（二手转述）｜R128 有术语收敛但始终没有成本坐标。Taskade 给出本库最直接一组（原始出处标注 Infragistics, July 2026，同一语料 100 条查询，RAG 管线 vs MCP 式 agentic retrieval）：成本 $0.0047→$0.1691（+3500%）；总延迟 12,715ms→31,932ms（+151%）；首 token 6,314ms→29,372ms（+365%）；答案胜率 38%→52%（+14 点）；每答引文数 2.2→4.5（+2.3）。它证明的是"agentic 检索买得到质量，但单价极高"，不是"agentic 检索应该被采用"——+14 点胜率对应约 +36 倍成本，净收益取决于单次回答的边际价值阈值（Taskade 判据："只在真正需要多步推理的查询上，质量的提升才值得成本；对简单事实型查询，agentic RAG 是纯浪费"）。引文数 +2.3 是被低估的一项：它正是 R128 归纳的"引证可核查"轴的工程侧证据。⚠️ 厂商博客二手转述，原始实验未直接核验、无代码、无样本量方法学说明、单一语料单一切分；只能作方向性对照，不能作任何成本模型的输入参数。

判断二 · 委托链权限断裂——284 的机制化解释｜R128 用 2610.30418 立了"安全分析单位从文档上移到结论"的开放 284。OvalEdge（2026-10-22）补工业侧机制：Agentic RAG 与标准 RAG 的差别是"人类不再在每一步触发检索并审阅输出"，故治理必须在运行时执行、作用于每一次检索（身份、权限、数据边界、可审计性）；引 Cloud Security Alliance 2026 称采用 NIST AI RMF 的企业中逾半数仍报多 agent 治理缺口（"agentic sprawl"）；失效模式是 revenue 分析 agent 把子查询委派给 admin 级权限的数据剖析 agent，剖析 agent 返回请求方分析师本不应看到的客户名与账户余额，根因是下游 agent 用了自己的权限而非发起请求用户的权限。因果链：权限判定的主体错了——不是"检索层没做权限过滤"（RelevanceLab 等通行做法恰恰写明"检索按请求用户的授权过滤，镜像源系统权限"），而是过滤做在了一层，但授权主体是 agent 而非用户；这解释了为什么 2610.30418 里"每份文档都合法"的聚合仍会越界——文档级过滤根本没失效，失效的是"谁的权限"在多 agent 链上没有唯一答案。RelevanceLab 与 R128 cyberhaven 表述一致，构成本轮一条可登记的工业共识：文档级权限过滤已是工程默认项，284 争点收窄到"复合结论级过滤在什么成本/召回代价下值得做"。⚠️ CSA 数据与失效模式均为厂商转述，未核到原始报告；"逾半数"未给样本量与调查口径。

判断三 · Reflective RAG 把"自评估"从方法论姿态变成训练目标｜Reflective RAG（ACL Findings 2026, Wu et al., 2026.findings-acl.648, pp.13252–13267）补上 R128 数据中心综述（开放 285）留下的一环——R128 已立"轨迹数据的 verify 阶段用什么自动判据"，因为轨迹正确性比单条回答正确性更难自动判定。路径是让模型自己产出判据：SFT 基于内部不确定性自校正标签学反思信号；RL 用反思优化策略 + 动态 KL 正则稳定训练；自述在 5 个知识密集 QA 基准上一致优于强 agentic 基线，多跳推理泛化更好、训练稳定性更好。① 部分解耦"评估语义"与"策略优化"（论文原话）——与 R126（2610.26260 重排评测须固定候选池）、R127（2610.28054 AutoSim 仿真器真实性、2610.27850 联合评估鲁棒性）追问的"我们有没有在正确地测量"直接对接：若反思信号本身可训练，则测量装置可以被优化，不再只是被批评的对象。② "内部不确定性"是承重墙——自评估信号的真实性完全依赖模型能区分"我确实会"与"我其实不会"。R128 立过"拒答率是错误的安全代理"（2610.30418：拒答本身可能暴露被禁结论），而 Reflective RAG 用了同一机制（模型自陈不确定）却得到正向用途——本库内部一处值得记录的表面矛盾：内部不确定性作为安全信号是错的，作为训练信号可能是对的。未经验证的判据假设：不确定性用来"决定说什么"时危险，用来做"决定练什么"时安全。⚠️ 只读到 abstract 层级，未读全文的实验设置、基准名称与数值、基线清单、RL 超参，不写入任何具体分数。

判断四 · 记忆策略与多租户权限是 repo_cards 里唯一有信息量的两处｜短期记忆三策略——删除（RemoveMessage）/ 修剪（trim_messages 按 last 策略、从 human 起、到 human 或 ai 止）/ 总结（summarize_conversation）；关键是工具消息排除（msg.type != 'tool'）与基于消息条数阈值的触发判定（count_messages），即压缩触发条件是对话长度而非语义重要性；持久层用 StateGraph + PostgresSaver checkpointer，即"对话状态"被建模为图状态而非简单 KV，与本库长期记忆线（236 共识的 MCP 协议化路径、2610.15715 NoteBank 线性注意力稀疏情景记忆）构成短/长两端连续设计。⚠️ 未经复现、版本与环境未核。devrochy/agent-core 的 per-tenant AgentSpec + Postgres RLS——把权限讨论从"检索时过滤"下推到存储层行级安全，RLS 由数据库强制执行，应用层漏写也无法绕过；与 Taskade/OvalEdge 构成三层权限模型：① 存储层 RLS ② 检索层按用户授权过滤 ③ 委托链授权主体（OvalEdge 指出此处断裂），三层由内到外、越外层越难强制。⚠️ 0⭐、experimental、卡片未给 RLS 策略样例。

工程侧｜工业侧词汇表已收敛且高度一致：规划 → 路由 → 迭代 → 校验（延续 R128）；框架支持面统一（LangGraph / LlamaIndex Agents / AutoGen / CrewAI）；模式清单稳定（Naive / Advanced / Modular / Agentic / GraphRAG / Adaptive / Vectorless；lyzr 的 Router / Agent-as-retriever / Corrective / Self-RAG / Adaptive / Agentic GraphRAG）。新增是成本口径：Smile 列四条常见错误（分块过大淹没相关信息、无管线评测致生产回答错误、索引不更新致结果过时、未做 GDPR 分析即云端 LLM 致注入即数据跨境转移）；cekura 给出数据集分工（训练/选检索器用 MS MARCO、NQ；比 embedding 用 BEIR、MTEB；测全管线用 HotpotQA、RGB、CRAG、FRAMES），本轮最实用的一句是"公开数据集适合挑零件，只有用你自己文档和真实查询建的测试集才能告诉你系统是否真的工作"；这与 R128 数据中心综述（开放 285）、R126/R127 评测批判同链：评测集来源必须是任务数据流水线，不是公开榜。⚠️ 全部是厂商/博客内容，无同行评审、无统一协议、无实验数字，只能作术语与关注点收敛的旁证。

3. 共识、争议与开放问题

共识 236（R129 续立并小幅扩写）｜Agent 长周期记忆基础设施协议化/本地化共识：R126 xerj 12498⭐ + CiteProof-AI MCP 审计工具包 + cortexkit/aft 语义搜索 IDE 层，与 R124 mcp-memory-service 2019⭐ + R120 shinpr/mcp-local-rag 415⭐ + R119 kordoc 2440⭐ + R124 slow-stack/mneme 160⭐ 构成"MCP 跨进程协议 + 本地优先数据主权"双路径收敛；R127/R128 无直接反证（iblai/os 走企业平台第三条路径）。R129 无直接反证，但 takeshy/obsidian-local-llm-hub（90⭐，含 MCP servers + RAG + 加密 + 编辑历史，简介明写 "Fully private, no cloud required"）是这条线的第 7 个同族样本，且是应用形态（app）而非框架——本地优先路径已从"框架/服务"下沉到"具体知识库应用的默认配置"；C 组短期记忆说明"对话状态 = 可检查点化的图状态"已被工业教程化。

争议 198（R129 续立）｜Agent 记忆基础设施 MCP 协议化 vs 私有协议：R126 新增对立面 CiteProof-AI（自有 MCP 审计工具包 + deterministic 引证核验）与 AgenticOS 类私有栈（vstorm-co/agenticos 声称 4700+ MCP servers 的自托管路线，🔍 待核）。R129 出现第三条材料：takeshy 同时列 "MCP servers" 与 "Fully private, no cloud required"——在一个应用里并置协议化与数据主权。不裁决任一方，但 198 需再次重述：真争点不是"协议 vs 私有"，而是"协议层被用来承载什么"——R128 给出第三个坐标（协议层可以只承担审计契约），R129 追加第四个（协议层与数据主权路径可并置而非互斥）。

开放 279（R129 增量）｜多 Agent VLM 冲突解决基准可验证性：R124-R125 CLASHarness 2610.19128 ground truth 依赖问题续立；R127 2610.27850、R128 2610.30418（禁止集比评分员更可核查）各为方法论参照。R129 新参照：Reflective RAG 的"基于内部不确定性自校正标签"SFT 阶段，本质是用模型自身信号构造标签，与 279 的"ground truth 从哪来"是同一问题的两种答案（人工枚举禁止集 vs 模型自校正）；禁止集可核查但覆盖有限，自校正可扩展但真伪不可核查——理想解可能需要两者结合。

开放 281（R129 增量）｜长上下文 vs 检索的边界不可迁移：R126 由 MMProLong 2605.13831（128K 无任务专属监督泛化）与 2610.24133（幂律遗忘 α≈0.55–0.7 / 0.2–0.5）给出理论两侧；R127 补工程侧成本（24K prompt 在 vLLM 触发 1.8–2.4s 调度冻结）；R128 补 Big Bird 稀疏注意力待查线索。R129 补第三个成本侧证据：agentic 检索总延迟 31,932ms、首 token 29,372ms——首 token / 总延迟比值压到 0.92，用户几乎要等完整检索链跑完才看到第一个字；三方（遗忘衰减 / 长窗口能力 / 推理排队）都有量化坐标，仍无跨模型可迁移判据。

开放 282（续立无增量）｜SID 映射稳定性与前缀平衡相容性：2610.26105（R126）主张动态场景下不重写 SID 映射以保语义邻域，2610.27898（R127）主张逐 token 解码阶段前缀平衡采样以保个性化；二者作用于同一对象的不同阶段，方向可能互相拉扯。R128/R129 无任何实验证据，仅作设计约束冲突登记。

开放 283（续立无增量）｜工业界"以可移植 Agent Skill 替换生产 RAG 栈"实证缺口：RecSys Vol.178 第 9 条 Airbnb 客服案例（🔍 正文未读到，无方法细节、无实测数字、无对照基线）。R129 无新证据；方向性张力未削弱——本轮 Smile、cekura、OvalEdge、Taskade、RelevanceLab 五份工业材料全部在加强 agentic 检索层（多步规划、权限过滤、评测集、成本管理），无一件主张移除 RAG 组件。

开放 284（R129 实质推进）｜RAG 安全的分析单位之争：逐文档授权 vs 复合结论级授权。R128 以 2610.30418 立此问题（单看任一文档均不越权、组合后泄漏复合结论；最强模型在良性单文档可答查询下仍有 7.08%（95% CI 3.75–10.67%）泄漏；加可见分类标签 + 显式聚合规则后跨全部意图 pooled 残余 unsafe disclosure 仍为 32.37%（95% CI 28.33–36.19%），两数口径不同不可相减）。R129 补工业侧机制解释：断裂点不在聚合层而在授权主体（OvalEdge：下游 agent 用了自己的权限而非用户的权限）；RelevanceLab/cyberhaven 一致表明"检索按请求用户授权过滤、镜像源系统权限"已是工程默认项。284 的问法再次收窄为"在多 agent 委托链上，权限主体如何随委托传递而保持不变"——技术上比"聚合后过滤"更根本，因为可在每一步委托上做。三处待补：① CSA 2026 报告原始数据（样本量、治理缺口定义）；② 是否有公开工作把"授权主体随委托传递"形式化；③ 2610.30418 的复合结论空间是否可枚举。

开放 285（R129 部分推进）｜Agentic RAG 数据构造：R128 以 ACL Findings 2026 数据中心综述立此问题（轨迹数据稀缺 + generate–verify–filter/refine 流水线），并问 ① verify 阶段用什么自动判据 ② filter/refine 代价如何随轨迹长度增长 ③ 跨领域任务难度是否可比。R129 回答了 ① 的一个候选答案：Reflective RAG 用"基于内部不确定性自校正标签"的 SFT 产出反思信号，再以 RL + 动态 KL 正则优化策略——即用模型自评估信号充当轨迹级判据；但自评估信号的可核查性未被解决（与 279 是同一问题两面）。② ③ 仍无证据：本轮无轨迹长度—代价消融实验，无跨领域难度可比性研究。⚠️ 只读到 abstract，故记为部分推进而非关闭。

趋势 343（R129 扩写：可核查性进入可训练时代）｜Agent-RAG 记忆/安全/冲突/自改进四维延展：R127 指出评测批判从单点走向成建制；R128 指出"可核查性"完成装置—对象—原料三环闭合，并分岔出权限可核查与引证可核查两条互不替代的轴。R129 的推进是第三条轴诞生：可核查性本身变成可训练目标。 Reflective RAG 不是又一件批评评测装置的工作（那属 R126/R127 装置批判线），而是把测量信号纳入优化回路——反思信号既是评估产物又是训练产物；这让"装置—对象—原料"三环之上多出一个回路：三环回答"测量是否可信、对象是否越界、原料从哪来"，回路回答"测量信号能否被优化"。四维落点：安全维从"复合结论级过滤 + 禁止集评测"具体化为"委托链上的授权主体传递"（OvalEdge 失效模式 + devrochy 存储层 RLS 补全三层模型）；冲突维（多 agent 聚合）第一次与安全维在机制层打通——聚合之所以危险，是因为聚合链条上权限主体发生了漂移；自改进维拿到最强增量（Reflective RAG）；记忆维由短期记忆三策略 + 本地优先 hub 续立。四维全部有实质增量，是 48 轮以来少见的四维齐涨。横切观察：成本从背景条件变成了被显式度量的对象——R128 记录工业侧只谈规划路由校验与权限、不谈代价，R129 把 +36 倍成本、+365% 首 token 延迟与胜率 +14 点并列在同一张表里；这与开放 283 构成对位：当 agentic 检索的成本被量化后，"用更轻的机制替代它"的经济动机就自动出现——283 的方向性张力第一次有了成本侧的量化前提。

4. State-of-the-Art 综述（R129 · 多轮深综合）

综述索引（R82-R129 累计 48 轮，80+ 件 RAG 主分类 net-new）：脉络见 §1-§3 与文末 ID Archive + changelog 累积。本轮 RAG 主分类 arXiv 净增 = 0；唯一新增 arXiv ID 2001.08361 为分类体系错位命中（主题: rag 但主分类 llm-infra，内容为预训练规模律），不计净增、不进 SOTA 主轴。另补入 1 件非 arXiv 学术材料：Reflective RAG（ACL Findings 2026, 2026.findings-acl.648）——本轮唯一具备训练方法论分量的新证据。

四轴 SOTA 主轴续立（R118 起不变，R129 无推翻）：① Is Agentic RAG Worth It? 2601.07711 ② Skill-RAG 2604.24594 ③ AgentSysBench 2608.15127 ④ Trustworthiness in RAG Systems 2409.10102 ⑤ AORCHESTRA 2602.03786 ⑥ Does Document Structure Help Dense Retrieval? 2610.10170 ⑦ The Geometry of Hierarchical Navigation 2610.12312 ⑧ Q-RAG 2511.07328 ⑨ Mapping the RAG Landscape 2610.01936。

R129 新增两项对照（均不构成主轴替换）：① Reflective RAG（自评估轴）——与 ⑨ Mapping the RAG Landscape 2610.01936（架构/方法全景）形成一对：后者是系统长什么样，前者是系统在训练时如何学会判断自己做得对不对；方法学结构：SFT 用内部不确定性自校正标签学"生成反思"→ RL 用反思优化策略 + 动态 KL 正则稳定；与 ① 2601.07711 形成张力——该文问"agentic RAG 值不值"，Reflective RAG 给出了"值不值取决于 agent 能否学会自我评估"这一可操作前提。② agentic 检索成本对照（工程轴）——Taskade 引 Infragistics 2026-07 的 100 查询对照，与 ⑥ 2610.10170、⑦ 2610.12312 这类检索质量轴工作形成对照：本轮这件讲的是买检索质量的单价。⚠️ 二手转述，单一语料，无方法学说明。

累计关键节点（48 轮）= R129 0 + R128 1 + R127 0 + R126 0 + R125 0 + R124 0 + R123 0 + R122 0 + R121 0 + R120 0 + R119 0 + R118 5 + R117 4 + R116 4 + R115 4 + R114 6 + R113 5 + R112-R82 累积 50+ 件（详见 ID Archive + changelog）。

5. 关键参考清单（R129 增量版 · 索引版）

R129 增量 8 件 evidence URL（完整清单见下方 curated 语料块）：A 组 5 repo_cards https://github.com/takeshy/obsidian-local-llm-hub· https://github.com/api-evangelist/pienso· https://github.com/Ablation-Tool/ablation· https://github.com/devrochy/agent-core· https://github.com/crmne/ruby_llm | B 组 Jay 博客动态 https://developer.ibm.com/articles/agentic-rag-pipeline· https://swyx.substack.com/p/claude-mods· https://github.com/NirDiamant/RAG_Techniques· https://github.com/PacktPublishing/Building-Production-Ready-AI-Agents· https://github.com/microsoft/ai-agent-runbooks· https://javarevisited.substack.com/p/30-essential-ai-engineering-concepts | C 组 Jay CSDN https://tianqi.csdn.net/6ad8b995536f457592b164b4.html· https://blog.csdn.net/weixin_29055699/article/details/168575246· https://machinelearningwithme.com/agent/chapter05/short-term-memory-rag-agent/ | D 组 1 paper_card https://arxiv.org/abs/2001.08361 | web_search 补 fresh https://aclanthology.org/2026.findings-acl.648· https://aclanthology.org/2026.findings-acl.648.pdf· https://doi.org/10.18653/v1/2026.findings-acl.648· https://www.taskade.com/blog/llm-rag-agent-agentic· https://www.ovaledge.com/blog/agentic-rag-governance· https://www.relevancelab.com/ai-services/rag-development· https://www.cekura.ai/discover/datasets-training-evaluating-rag-models· https://smile.eu/en/node/878· https://onepixelsoft.com/blog-details/rag-in-2026-evolution-future-of-ai-retrieval

R128 增量 8 件 evidence URL：A 组 3 repo_cards https://github.com/parthchandak02/literature-review-assistant· https://github.com/YenochQin/llm-wiki· https://github.com/Aryan-Pardeshi/CiteProof-AI-MCP | B 组 ByteByteGo https://blog.bytebytego.com/p/ep230-rag-vs-agentic-rag | C 组 Tom 雷达 https://arxiv.org/abs/2610.30418· https://www.trydeepteam.com/guides/guide-red-teaming-agentic-rag· https://www.redfoxsec.com/blog/rag-pipeline-attack-scenarios-context-manipulation-and-data-leakage-explained | D 组 3 paper_cards https://arxiv.org/abs/2301.12597· https://arxiv.org/abs/1506.06726· https://arxiv.org/abs/2007.14062 | web_search 补 fresh https://aclanthology.org/2026.findings-acl.78· https://aclanthology.org/2026.findings-acl.78.pdf

续立 URL（R127-R117）：R127 https://recsys.substack.com/i/220663591/9-frustratingly-easy-agentic-rag-via-portable-skills-a-case-study-in-customer-support | R126 https://github.com/yuxiaopeng/Github-Ranking-AI/blob/main/Top100/AI%20Agents.md | R124 https://arxiv.org/abs/2610.12085 https://tianqi.csdn.net/6ac89a94ad85947e68cebcf1.html | R121 https://arxiv.org/abs/2603.15031 https://arxiv.org/abs/2606.07001 | R120 https://arxiv.org/abs/2501.09136v4 https://arxiv.org/abs/2602.05758 https://arxiv.org/abs/2610.18092 https://arxiv.org/abs/2610.18579 https://papers.cool/arxiv/2610.18410 | R117 https://arxiv.org/abs/2610.01936 https://arxiv.org/abs/2610.10170 https://arxiv.org/abs/2610.12312

R129 arXiv ID 增量（1 件）：2001.08361 Scaling Laws（主题: rag 命中但主分类 llm-infra，分类体系错位命中，不计主分类净增）｜ R128 arXiv ID 增量（4 件）：2610.30418 Agent Mosaic（净增 1）+ 2301.12597 BLIP-2 + 1506.06726 Skip-Thought Vectors + 2007.14062 Big Bird（后 3 件 主题: rag 错位命中，主分类 engineering / llm-infra / llm-infra，不计净增）｜ R127 arXiv ID 增量（5 件）：见 §2 B 组，全部 cs.IR 邻接级，净增 = 0 ｜ R126 arXiv ID 增量（9 件）：MemT-1.0 / KronLSTM / 条件无毛定理 / OPSD / AGR / NevIR / Babbar26 / FLMNet / SID 拓扑保持｜ R125 arXiv ID 增量（4 件）：2610.09872 LiveMACE + 2610.16402 EvoWeb + 2610.21082 SciΠvot + 2610.23194 OfficeTown ｜ R124-R82 arXiv ID 增量（R82-R120 293+ 件 + R124-R121 14 件，完整清单见下方 arxiv 语料块；仅此 4 件未在 URL 语料中以裸 ID 出现）：2610.02188 2610.12461 2610.17814 2610.18818。

CVE 14：CVE-2025-68664, CVE-2026-20805, CVE-2026-21858, CVE-2026-22778, CVE-2026-26030, CVE-2026-3172, CVE-2026-45829, CVE-2026-4810, CVE-2026-49468, CVE-2026-50548, CVE-2026-50549, CVE-2026-54309, CVE-2026-55255, CVE-2026-56274。

DOI 41：10.48550/arxiv.2001.09977 10.48550/arxiv.2009.14794 10.48550/arxiv.2610.12085 10.48550/arxiv.2610.13695 10.48550/arxiv.2610.15715 10.48550/arxiv.2610.19128 10.48550/arxiv.2610.20431 10.48550/arxiv.2610.13623 10.48550/arxiv.2610.20939 10.48550/arxiv.2610.22276 10.48550/arxiv.2312.10997 10.48550/arxiv.2409.10102 10.48550/arxiv.2511.07328 10.48550/arxiv.2511.02230 10.48550/arxiv.2601.07711 10.48550/arxiv.2602.03786 10.48550/arxiv.2604.24594 10.48550/arxiv.2605.12357 10.48550/arxiv.2605.13831 10.48550/arxiv.2606.06090 10.48550/arxiv.2606.14061 10.48550/arxiv.2606.26511 10.48550/arxiv.2606.27708 10.48550/arxiv.2607.01224 10.48550/arxiv.2607.01935 10.48550/arxiv.2607.26760 10.48550/arxiv.2608.15127 10.48550/arxiv.2609.01836 10.48550/arxiv.2609.05339 10.48550/arxiv.2609.39075 10.48550/arxiv.2609.39096 10.48550/arxiv.2610.00972 10.48550/arxiv.2610.01509 10.48550/arxiv.2610.01936 10.48550/arxiv.2610.02150 10.48550/arxiv.2610.10170 10.48550/arxiv.2610.10515 10.48550/arxiv.2610.11899 10.48550/arxiv.2610.12085 10.48550/arxiv.2610.12312 10.48550/arxiv.2610.12415 10.48550/arxiv.2510.09665 10.48550/arxiv.2610.30418 10.48550/arxiv.2301.12597 10.48550/arxiv.1506.06726 10.48550/arxiv.2007.14062 R129 新增 2 件：10.48550/arxiv.2001.08361 10.18653/v1/2026.findings-acl.648。DOI（含 URL）：https://doi.org/10.48550/arxiv.2009.14794 https://doi.org/10.48550/arxiv.2610.12085 https://doi.org/10.48550/arxiv.2610.13695 https://doi.org/10.18653/v1/2026.findings-acl.648 10.48550/arxiv.2009.14794 10.48550/arxiv.2610.13695 10.48550/arxiv.2610.22280。

R129/R128 新增来源 URL：已并入下方 curated 语料块（arxiv / github / vllm / huggingface / csdn / substack 六组），不重复列出。 URL (corpus curation required to maintain missing=0):

arxiv (491) http://arxiv.org/abs/2409.10102v2 http://arxiv.org/abs/2511.07328v2 http://arxiv.org/abs/2601.07711v2 http://arxiv.org/abs/2602.03786v1 http://arxiv.org/abs/2604.24594v1 http://arxiv.org/abs/2606.26511v1 http://arxiv.org/abs/2607.01935v1 http://arxiv.org/abs/2608.15127v1 http://arxiv.org/abs/2608.21252v1 http://arxiv.org/abs/2608.22752v1 http://arxiv.org/abs/2608.22872v1 http://arxiv.org/abs/2609.37673v1 http://arxiv.org/abs/2609.37749v1 http://arxiv.org/abs/2609.39075v1 http://arxiv.org/abs/2609.39929v1 http://arxiv.org/abs/2610.00972 http://arxiv.org/abs/2610.01509 http://arxiv.org/abs/2610.01767v1 http://arxiv.org/abs/2610.01936v1 http://arxiv.org/abs/2610.05782v1 http://arxiv.org/abs/2610.05826v1 http://arxiv.org/abs/2610.06207v1 http://arxiv.org/abs/2610.08452v1 http://arxiv.org/abs/2610.08463v1 http://arxiv.org/abs/2610.08571v1 http://arxiv.org/abs/2610.08674v1 http://arxiv.org/abs/2610.10091v1 http://arxiv.org/abs/2610.10170v1 http://arxiv.org/abs/2610.10507v1 http://arxiv.org/abs/2610.11169v1 http://arxiv.org/abs/2610.11899v1 http://arxiv.org/abs/2610.12085v1 http://arxiv.org/abs/2610.12312v1 http://arxiv.org/abs/2610.12415v1 https://arxiv.org/abs/2005.11401 https://arxiv.org/abs/2312.10997 https://arxiv.org/abs/2401.00396 https://arxiv.org/abs/2402.07867 https://arxiv.org/abs/2409.01666 https://arxiv.org/abs/2410.05983 https://arxiv.org/abs/2412.11854 https://arxiv.org/abs/2501.09136 https://arxiv.org/abs/2502.08826 https://arxiv.org/abs/2504.01840 https://arxiv.org/abs/2504.09554 https://arxiv.org/abs/2505.23052 https://arxiv.org/abs/2506.03401 https://arxiv.org/abs/2506.05176 https://arxiv.org/abs/2507.02962 https://arxiv.org/abs/2507.05257 https://arxiv.org/abs/2509.12541 https://arxiv.org/abs/2510.15253 https://arxiv.org/abs/2512.09487 https://arxiv.org/abs/2512.09695v4 https://arxiv.org/abs/2512.14629 https://arxiv.org/abs/2601.02993 https://arxiv.org/abs/2601.07504 https://arxiv.org/abs/2601.07978 https://arxiv.org/abs/2601.12538 https://arxiv.org/abs/2601.23254 https://arxiv.org/abs/2602.00296 https://arxiv.org/abs/2602.03442 https://arxiv.org/abs/2602.03992 https://arxiv.org/abs/2602.06052 https://arxiv.org/abs/2602.08005 https://arxiv.org/abs/2602.11671 https://arxiv.org/abs/2603.00873 https://arxiv.org/abs/2603.07379 https://arxiv.org/abs/2603.25152 https://arxiv.org/abs/2604.04359 https://arxiv.org/abs/2604.08224 https://arxiv.org/abs/2604.09666 https://arxiv.org/abs/2604.14572 https://arxiv.org/abs/2605.01495 https://arxiv.org/abs/2605.01970v4 https://arxiv.org/abs/2605.12357 https://arxiv.org/abs/2605.16352 https://arxiv.org/abs/2605.21071v4 https://arxiv.org/abs/2605.27123 https://arxiv.org/abs/2605.28120 https://arxiv.org/abs/2605.29640v1 https://arxiv.org/abs/2606.00610v1 https://arxiv.org/abs/2606.01613 https://arxiv.org/abs/2606.10953 https://arxiv.org/abs/2606.16817 https://arxiv.org/abs/2606.16903 https://arxiv.org/abs/2606.18037 https://arxiv.org/abs/2606.26458 https://arxiv.org/abs/2606.26916 https://arxiv.org/abs/2606.27708 https://arxiv.org/abs/2607.01224 https://arxiv.org/abs/2607.08716 https://arxiv.org/abs/2607.17715 https://arxiv.org/abs/2607.18152 https://arxiv.org/abs/2607.21596 https://arxiv.org/abs/2607.22042 https://arxiv.org/abs/2607.24748 https://arxiv.org/abs/2607.26497v2 https://arxiv.org/abs/2607.26760 https://arxiv.org/abs/2607.29677 https://arxiv.org/abs/2608.00902 https://arxiv.org/abs/2608.01526 https://arxiv.org/abs/2608.02870 https://arxiv.org/abs/2608.08340 https://arxiv.org/abs/2608.08466 https://arxiv.org/abs/2608.09802 https://arxiv.org/abs/2608.11030 https://arxiv.org/abs/2608.11660 https://arxiv.org/abs/2608.12218 https://arxiv.org/abs/2608.12253 https://arxiv.org/abs/2608.12304 https://arxiv.org/abs/2608.12875 https://arxiv.org/abs/2608.13010 https://arxiv.org/abs/2608.13120 https://arxiv.org/abs/2608.13160 https://arxiv.org/abs/2608.13237 https://arxiv.org/abs/2608.13410 https://arxiv.org/abs/2608.13545 https://arxiv.org/abs/2608.13547 https://arxiv.org/abs/2608.14036 https://arxiv.org/abs/2608.14054 https://arxiv.org/abs/2608.14210 https://arxiv.org/abs/2608.14229 https://arxiv.org/abs/2608.15767 https://arxiv.org/abs/2608.15888 https://arxiv.org/abs/2608.15994 https://arxiv.org/abs/2608.16490 https://arxiv.org/abs/2608.16515 https://arxiv.org/abs/2608.16536 https://arxiv.org/abs/2608.16628 https://arxiv.org/abs/2608.16776 https://arxiv.org/abs/2608.16885 https://arxiv.org/abs/2608.17050 https://arxiv.org/abs/2608.17536 https://arxiv.org/abs/2608.17744 https://arxiv.org/abs/2608.17781 https://arxiv.org/abs/2608.17950v1 https://arxiv.org/abs/2608.17960 https://arxiv.org/abs/2608.18184 https://arxiv.org/abs/2608.18271 https://arxiv.org/abs/2608.18489 https://arxiv.org/abs/2608.18607 https://arxiv.org/abs/2608.18613 https://arxiv.org/abs/2608.18731 https://arxiv.org/abs/2608.18852 https://arxiv.org/abs/2608.19269 https://arxiv.org/abs/2608.19758 https://arxiv.org/abs/2608.19799 https://arxiv.org/abs/2608.19857 https://arxiv.org/abs/2608.20246 https://arxiv.org/abs/2608.20281 https://arxiv.org/abs/2608.20317 https://arxiv.org/abs/2608.20335 https://arxiv.org/abs/2608.21450 https://arxiv.org/abs/2608.23252 https://arxiv.org/abs/2608.24040 https://arxiv.org/abs/2608.24053 https://arxiv.org/abs/2608.25500 https://arxiv.org/abs/2608.25625 https://arxiv.org/abs/2608.25735 https://arxiv.org/abs/2608.25757 https://arxiv.org/abs/2608.25986 https://arxiv.org/abs/2608.26091 https://arxiv.org/abs/2608.26530 https://arxiv.org/abs/2608.26623 https://arxiv.org/abs/2608.26836 https://arxiv.org/abs/2608.27448 https://arxiv.org/abs/2608.27455 https://arxiv.org/abs/2608.27809 https://arxiv.org/abs/2608.28389 https://arxiv.org/abs/2608.31082 https://arxiv.org/abs/2608.31139 https://arxiv.org/abs/2609.01343 https://arxiv.org/abs/2609.01601 https://arxiv.org/abs/2609.01657 https://arxiv.org/abs/2609.01836 https://arxiv.org/abs/2609.03005 https://arxiv.org/abs/2609.03712 https://arxiv.org/abs/2609.03756 https://arxiv.org/abs/2609.05110 https://arxiv.org/abs/2609.05339v1 https://arxiv.org/abs/2609.05736 https://arxiv.org/abs/2609.06702 https://arxiv.org/abs/2609.08126 https://arxiv.org/abs/2609.08365 https://arxiv.org/abs/2609.08572 https://arxiv.org/abs/2609.08832 https://arxiv.org/abs/2609.09875 https://arxiv.org/abs/2609.10539 https://arxiv.org/abs/2609.10745 https://arxiv.org/abs/2609.11209 https://arxiv.org/abs/2609.11294 https://arxiv.org/abs/2609.11390 https://arxiv.org/abs/2609.11561 https://arxiv.org/abs/2609.11572 https://arxiv.org/abs/2609.11646 https://arxiv.org/abs/2609.11808 https://arxiv.org/abs/2609.13463 https://arxiv.org/abs/2609.14302 https://arxiv.org/abs/2609.14320 https://arxiv.org/abs/2609.15126 https://arxiv.org/abs/2609.15578 https://arxiv.org/abs/2609.15800 https://arxiv.org/abs/2609.15830 https://arxiv.org/abs/2609.15991 https://arxiv.org/abs/2609.16409 https://arxiv.org/abs/2609.16591 https://arxiv.org/abs/2609.16816 https://arxiv.org/abs/2609.16818 https://arxiv.org/abs/2609.17012 https://arxiv.org/abs/2609.17320 https://arxiv.org/abs/2609.17652 https://arxiv.org/abs/2609.17653 https://arxiv.org/abs/2609.18063 https://arxiv.org/abs/2609.19656 https://arxiv.org/abs/2609.19969 https://arxiv.org/abs/2609.20423 https://arxiv.org/abs/2609.20715 https://arxiv.org/abs/2609.22086 https://arxiv.org/abs/2609.22319 https://arxiv.org/abs/2609.22712 https://arxiv.org/abs/2609.23551 https://arxiv.org/abs/2609.24220 https://arxiv.org/abs/2609.25053 https://arxiv.org/abs/2609.25267 https://arxiv.org/abs/2609.25636 https://arxiv.org/abs/2609.25804 https://arxiv.org/abs/2609.25853 https://arxiv.org/abs/2609.25991 https://arxiv.org/abs/2609.26489 https://arxiv.org/abs/2609.26774 https://arxiv.org/abs/2609.26781 https://arxiv.org/abs/2609.27213 https://arxiv.org/abs/2609.29845 https://arxiv.org/abs/2609.30192 https://arxiv.org/abs/2609.30904 https://arxiv.org/abs/2609.31002 https://arxiv.org/abs/2609.31415 https://arxiv.org/abs/2609.32049 https://arxiv.org/abs/2609.32259 https://arxiv.org/abs/2609.32722 https://arxiv.org/abs/2609.33403 https://arxiv.org/abs/2609.33485 https://arxiv.org/abs/2609.33987 https://arxiv.org/abs/2609.34227 https://arxiv.org/abs/2609.34242 https://arxiv.org/abs/2609.34385 https://arxiv.org/abs/2609.34392 https://arxiv.org/abs/2609.35427 https://arxiv.org/abs/2609.35530 https://arxiv.org/abs/2609.35629 https://arxiv.org/abs/2609.36138 https://arxiv.org/abs/2609.36314 https://arxiv.org/abs/2609.36322 https://arxiv.org/abs/2609.36333 https://arxiv.org/abs/2609.36636 https://arxiv.org/abs/2609.36730 https://arxiv.org/abs/2609.37147 https://arxiv.org/abs/2609.37559 https://arxiv.org/abs/2609.37687 https://arxiv.org/abs/2609.37690 https://arxiv.org/abs/2609.37863 https://arxiv.org/abs/2609.38270v1 https://arxiv.org/abs/2609.38349 https://arxiv.org/abs/2609.38660 https://arxiv.org/abs/2609.38721 https://arxiv.org/abs/2609.39102 https://arxiv.org/abs/2609.39154 https://arxiv.org/abs/2609.39788 https://arxiv.org/abs/2609.39841 https://arxiv.org/abs/2609.39903 https://arxiv.org/abs/2609.39982 https://arxiv.org/abs/2609.40222 https://arxiv.org/abs/2609.40316 https://arxiv.org/abs/2610.00437 https://arxiv.org/abs/2610.00544 https://arxiv.org/abs/2610.00559 https://arxiv.org/abs/2610.00812 https://arxiv.org/abs/2610.01027v1 https://arxiv.org/abs/2610.01092 https://arxiv.org/abs/2610.01428 https://arxiv.org/abs/2610.01871 https://arxiv.org/abs/2610.02002v1 https://arxiv.org/abs/2610.02150 https://arxiv.org/abs/2610.02196 https://arxiv.org/abs/2610.03136v1 https://arxiv.org/abs/2610.03421v1 https://arxiv.org/abs/2610.03632v1 https://arxiv.org/abs/2610.05622 https://arxiv.org/abs/2610.07332 https://arxiv.org/abs/2610.08244 https://arxiv.org/abs/2610.13623 https://arxiv.org/abs/2610.15715 https://arxiv.org/abs/2610.17901 https://arxiv.org/abs/2610.19128 https://arxiv.org/abs/2610.20431 https://arxiv.org/abs/2610.20578 https://arxiv.org/abs/2610.20939 https://arxiv.org/abs/2610.21220 https://arxiv.org/abs/2610.21330 https://arxiv.org/abs/2610.21877 https://arxiv.org/abs/2610.21893 https://arxiv.org/abs/2610.22200 https://arxiv.org/abs/2610.22276 https://arxiv.org/html/2510.11541v2 https://arxiv.org/html/2603.10765v1 https://arxiv.org/abs/2001.09977 https://arxiv.org/abs/2606.14061 https://arxiv.org/abs/2510.09665 https://arxiv.org/abs/2511.02230 https://arxiv.org/abs/2009.14794 https://arxiv.org/abs/2610.13695 https://arxiv.org/abs/2610.22280 https://arxiv.org/abs/2610.09872 https://arxiv.org/abs/2610.23194 https://arxiv.org/abs/2610.21082 https://arxiv.org/abs/2610.16402 https://arxiv.org/abs/2610.10515 https://arxiv.org/abs/2610.26825 https://arxiv.org/abs/2610.26834 https://arxiv.org/abs/2610.24133 https://arxiv.org/abs/2610.22512 https://arxiv.org/abs/2605.13831 https://arxiv.org/abs/2606.06090 https://arxiv.org/abs/2409.10102 https://arxiv.org/abs/2510.15253v3 https://arxiv.org/abs/2511.02230v6 https://arxiv.org/abs/2511.07328 https://arxiv.org/abs/2601.07711 https://arxiv.org/abs/2602.03786 https://arxiv.org/abs/2603.00873v1 https://arxiv.org/abs/2603.07379v1 https://arxiv.org/abs/2604.04359v1 https://arxiv.org/abs/2604.24594 https://arxiv.org/abs/2605.12357v2 https://arxiv.org/abs/2606.16817v1 https://arxiv.org/abs/2606.26511 https://arxiv.org/abs/2607.01935 https://arxiv.org/abs/2607.17715v1 https://arxiv.org/abs/2607.26760v1 https://arxiv.org/abs/2608.00902v1 https://arxiv.org/abs/2608.01526v1 https://arxiv.org/abs/2608.12304v1 https://arxiv.org/abs/2608.12304v1v1 https://arxiv.org/abs/2608.12875v1 https://arxiv.org/abs/2608.13010v1 https://arxiv.org/abs/2608.13160v1 https://arxiv.org/abs/2608.13237v1 https://arxiv.org/abs/2608.13410v1 https://arxiv.org/abs/2608.14054v1 https://arxiv.org/abs/2608.14210v1 https://arxiv.org/abs/2608.15127 https://arxiv.org/abs/2608.16515v1 https://arxiv.org/abs/2608.16536v1 https://arxiv.org/abs/2608.16628v1 https://arxiv.org/abs/2608.16776v1 https://arxiv.org/abs/2608.17536v1 https://arxiv.org/abs/2608.17781v1 https://arxiv.org/abs/2608.17960v1 https://arxiv.org/abs/2608.18184v1 https://arxiv.org/abs/2608.18489v1 https://arxiv.org/abs/2608.18607v1 https://arxiv.org/abs/2608.18613v1 https://arxiv.org/abs/2608.19857v1 https://arxiv.org/abs/2608.20246v1 https://arxiv.org/abs/2608.20281v1 https://arxiv.org/abs/2608.21252 https://arxiv.org/abs/2608.21252v1 https://arxiv.org/abs/2608.22752 https://arxiv.org/abs/2608.22752v1 https://arxiv.org/abs/2608.22872 https://arxiv.org/abs/2608.23252v1 https://arxiv.org/abs/2608.24040v1 https://arxiv.org/abs/2608.25500v1 https://arxiv.org/abs/2608.25625v1 https://arxiv.org/abs/2608.25986v1 https://arxiv.org/abs/2608.26091v1 https://arxiv.org/abs/2608.26530v1 https://arxiv.org/abs/2608.26836v1 https://arxiv.org/abs/2608.27448v1 https://arxiv.org/abs/2608.27455v1 https://arxiv.org/abs/2608.27809v1 https://arxiv.org/abs/2608.28389v1 https://arxiv.org/abs/2609.37673 https://arxiv.org/abs/2609.37749 https://arxiv.org/abs/2609.39075 https://arxiv.org/abs/2609.39929 https://arxiv.org/abs/2610.00972 https://arxiv.org/abs/2610.01509 https://arxiv.org/abs/2610.01767 https://arxiv.org/abs/2610.01871v1 https://arxiv.org/abs/2610.05782v1 https://arxiv.org/abs/2610.05826v1 https://arxiv.org/abs/2610.06207v1 https://arxiv.org/abs/2610.08452v1 https://arxiv.org/abs/2610.08463v1 https://arxiv.org/abs/2610.08571v1 https://arxiv.org/abs/2610.08674v1 https://arxiv.org/abs/2610.10091v1 https://arxiv.org/abs/2610.10170v1 https://arxiv.org/abs/2610.10507v1 https://arxiv.org/abs/2610.11169v1 https://arxiv.org/abs/2610.11899v1 https://arxiv.org/abs/2610.12085v1 https://arxiv.org/abs/2610.13695v1 https://arxiv.org/html/2409.10102v2 https://arxiv.org/html/2601.07978 https://arxiv.org/html/2604.09666v1 https://arxiv.org/html/2604.14572v3 https://arxiv.org/html/2607.24748 https://arxiv.org/html/2607.24748v1 https://arxiv.org/html/2608.21252v1 https://arxiv.org/html/2609.25853 https://arxiv.org/html/2609.25991v1 https://arxiv.org/pdf/2312.10997 https://arxiv.org/pdf/2409.10102 https://arxiv.org/pdf/2510.09665 https://arxiv.org/pdf/2602.03442 https://arxiv.org/pdf/2603.07379 https://arxiv.org/pdf/2605.12357 https://arxiv.org/pdf/2607.17715 https://arxiv.org/pdf/2607.26760 https://arxiv.org/pdf/2607.29677 https://arxiv.org/pdf/2608.00902 https://arxiv.org/pdf/2608.01526 https://arxiv.org/pdf/2608.09802 https://arxiv.org/pdf/2608.12304 https://arxiv.org/pdf/2608.13010 https://arxiv.org/pdf/2608.13237 https://arxiv.org/pdf/2608.13410 https://arxiv.org/pdf/2608.18184 https://arxiv.org/pdf/2608.18489 https://arxiv.org/pdf/2608.19269 https://arxiv.org/pdf/2608.19857 https://arxiv.org/pdf/2608.21252 https://arxiv.org/pdf/2608.22752 https://arxiv.org/pdf/2608.23252 https://arxiv.org/pdf/2608.25500 https://arxiv.org/pdf/2608.25625 https://arxiv.org/pdf/2608.25735 https://arxiv.org/pdf/2608.25757 https://arxiv.org/pdf/2608.25986 https://arxiv.org/pdf/2608.26091 https://arxiv.org/pdf/2608.26530 https://arxiv.org/pdf/2608.26836 https://arxiv.org/pdf/2608.27448 https://arxiv.org/pdf/2608.27455 https://arxiv.org/pdf/2608.27809 https://arxiv.org/pdf/2608.28389 https://arxiv.org/pdf/2608.31139 https://arxiv.org/pdf/2609.01343 https://arxiv.org/pdf/2609.01601 https://arxiv.org/pdf/2609.01836 https://arxiv.org/pdf/2610.10170 https://arxiv.org/pdf/2610.12312 https://arxiv.org/abs/2610.30418 https://arxiv.org/abs/2301.12597 https://arxiv.org/abs/1506.06726 https://arxiv.org/abs/2007.14062 https://arxiv.org/abs/2511.07328v2 https://arxiv.org/abs/2610.12312v1 https://arxiv.org/abs/2610.12415v1 papers.cool (23) https://papers.cool/cs.IR/feed https://papers.cool/arxiv/cs.IR/feed https://papers.cool/arxiv/2609.05110 https://papers.cool/arxiv/2610.17823 https://papers.cool/arxiv/2610.17841 https://papers.cool/arxiv/2610.18622 https://papers.cool/arxiv/2609.25991 https://papers.cool/arxiv/2610.19578 https://papers.cool/arxiv/2610.19608 https://papers.cool/arxiv/2610.20048 https://papers.cool/arxiv/2610.20122 https://papers.cool/arxiv/2610.20578 https://papers.cool/arxiv/2610.21220 https://papers.cool/arxiv/2610.21330 https://papers.cool/arxiv/2610.21877 https://papers.cool/arxiv/2610.21893 https://papers.cool/arxiv/2610.21030 https://papers.cool/arxiv/2610.26831 https://papers.cool/arxiv/2610.26260 https://papers.cool/arxiv/2610.26220 https://papers.cool/arxiv/2610.26143 https://papers.cool/arxiv/2610.26105 https://papers.cool/arxiv/2610.28569 https://papers.cool/arxiv/2610.28435 https://papers.cool/arxiv/2610.28054 https://papers.cool/arxiv/2610.27898 https://papers.cool/arxiv/2610.27850 http://arxiv.org/abs/2001.08361v1

github (75) https://github.com/Ayanami0730/arag https://github.com/DavidZWZ/Awesome-RAG-Reasoning https://github.com/OpenBMB/UltraRAG https://github.com/ParticleMedia/RAGTruth https://github.com/RUC-NLPIR/Awesome-Long-Horizon-Agents https://github.com/Shubhamsaboo/awesome-llm-apps https://github.com/Tencent/WeMM-Embedding https://github.com/YennNing/MC-Search https://github.com/YunjiaXi/Awesome-Search-Agent-Papers https://github.com/abdoslamb/inkdoc https://github.com/aitherium/awresearch https://github.com/apowerb/apowerb https://github.com/chrisryugj/kordoc https://github.com/doobidoo/mcp-memory-service https://github.com/dukesun99/Corpus2Skill https://github.com/edinburghnlp/mmlongbench https://github.com/elizaos/eliza https://github.com/future-agi/future-agi https://github.com/google-research/veriharness https://github.com/griver/Q-RAG https://github.com/langchain-ai/langgraph/releases https://github.com/liaojunchen/arxiv-daily https://github.com/llm-d/llm-d https://github.com/llm-lab-org/multimodal-rag-survey https://github.com/mattral/rag-multimodal-financial-doc-analysis-and-recall https://github.com/modelcontextprotocol https://github.com/neulab/think-before-you-link https://github.com/odakin/arxiv-digest https://github.com/omkarbhad/agentforge https://github.com/piterkai/nova-retrieve https://github.com/run-llama/ExtractBench https://github.com/run-llama/llama_index https://github.com/senna-lang/arxiv-compass https://github.com/topics/generative-ai-agents https://github.com/topics/production-ai https://github.com/umwyf/CRITICL https://github.com/volcengine/OpenViking https://github.com/weitianxin/Awesome-Agentic-Reasoning https://github.com/yzhao062/awesome-auditable-ai https://github.com/zilliztech/deep-searcher https://github.com/ziqiwang0908/RAGRouter-Bench https://github.com/AbdoslamB/InkDoc https://github.com/Aitherium/awresearch https://github.com/LiaojunChen/Arxiv-daily https://github.com/Mattral/RAG-Multimodal-Financial-Doc-Analysis-and-Recall https://github.com/Harmeet10000/AgentNexus-LangChain-FastAPI https://github.com/semrincaballero/Polymarket-Prediction-Bot https://github.com/PersonalClaw/PersonalClaw https://github.com/slow-stack/mneme https://github.com/xerj-org/xerj https://github.com/ai-boost/awesome-harness-engineering/blob/main/README.md https://github.com/ather-techie/rag-interview-system https://github.com/microsoft/agent-framework-go https://github.com/khaledjamal/ai-foundry-workshop https://github.com/AdilShamim8/GenAI-Roadmap-with-Notes-and-Projects https://github.com/metadist/synaplan https://github.com/LapusneanuNicolae/veritas-rag https://github.com/ram-singh1/parallel-rag-agent https://github.com/shinpr/mcp-local-rag https://github.com/Aryan-Pardeshi/CiteProof-AI https://github.com/cortexkit/aft https://github.com/HyxiaoGe/fusion https://github.com/worldbench/awesome-ai-auto-research https://github.com/iblai/os https://github.com/parthchandak02/literature-review-assistant https://github.com/YenochQin/llm-wiki https://github.com/Aryan-Pardeshi/CiteProof-AI-MCP https://github.com/takeshy/obsidian-local-llm-hub https://github.com/api-evangelist/pienso https://github.com/Ablation-Tool/ablation https://github.com/devrochy/agent-core https://github.com/crmne/ruby_llm https://github.com/NirDiamant/RAG_Techniques https://github.com/PacktPublishing/Building-Production-Ready-AI-Agents https://github.com/microsoft/ai-agent-runbooks

vllm (3) https://github.com/vllm-project/vllm/issues/61007 https://github.com/vllm-project/vllm/actions/runs/38025534578 https://github.com/vllm-project/vllm/actions/runs/37362806856

huggingface (83) https://huggingface.co/Qwen/Qwen3-Embedding-8B https://huggingface.co/blog/hotchpotch/mlateon-multilingual-colbert-hakari-bench https://huggingface.co/blog/huggingface/state-of-os-hf-spring-2026 https://huggingface.co/blog/state-of-open-models-summer-2026 https://huggingface.co/cohere/rerank-v3.5 https://huggingface.co/datasets/neulab/merlin-rare https://huggingface.co/datasets/paiteq-ai/rag-bench-2026q2 https://huggingface.co/nvidia/nemotron-colembed-vl-8b-v2 https://huggingface.co/papers/2312.10997 https://huggingface.co/papers/2409.10102 https://huggingface.co/papers/2501.09136 https://huggingface.co/papers/2511.07328 https://huggingface.co/papers/2601.07504 https://huggingface.co/papers/2601.07711 https://huggingface.co/papers/2602.03442 https://huggingface.co/papers/2602.03786 https://huggingface.co/papers/2603.07379 https://huggingface.co/papers/2604.14572 https://huggingface.co/papers/2604.24594 https://huggingface.co/papers/2605.12357 https://huggingface.co/papers/2606.16817 https://huggingface.co/papers/2607.26760 https://huggingface.co/papers/2607.29677 https://huggingface.co/papers/2608.12304 https://huggingface.co/papers/2608.12875 https://huggingface.co/papers/2608.13010 https://huggingface.co/papers/2608.13237 https://huggingface.co/papers/2608.13410 https://huggingface.co/papers/2608.15127 https://huggingface.co/papers/2608.16515 https://huggingface.co/papers/2608.16536 https://huggingface.co/papers/2608.16628 https://huggingface.co/papers/2608.16776 https://huggingface.co/papers/2608.17050 https://huggingface.co/papers/2608.17536 https://huggingface.co/papers/2608.17781 https://huggingface.co/papers/2608.17960 https://huggingface.co/papers/2608.18184 https://huggingface.co/papers/2608.18489 https://huggingface.co/papers/2608.18607 https://huggingface.co/papers/2608.18613 https://huggingface.co/papers/2608.19857 https://huggingface.co/papers/2608.20246 https://huggingface.co/papers/2608.20281 https://huggingface.co/papers/2608.21252 https://huggingface.co/papers/2608.22752 https://huggingface.co/papers/2608.23252 https://huggingface.co/papers/2608.25500 https://huggingface.co/papers/2608.25625 https://huggingface.co/papers/2608.25735 https://huggingface.co/papers/2608.25986 https://huggingface.co/papers/2608.26091 https://huggingface.co/papers/2608.26530 https://huggingface.co/papers/2608.26623 https://huggingface.co/papers/2608.26836 https://huggingface.co/papers/2608.27448 https://huggingface.co/papers/2608.27455 https://huggingface.co/papers/2608.27809 https://huggingface.co/papers/2608.28389 https://huggingface.co/papers/2608.31082 https://huggingface.co/papers/2608.31139 https://huggingface.co/papers/2609.01343 https://huggingface.co/papers/2609.01601 https://huggingface.co/papers/2609.01836 https://huggingface.co/papers/2609.05339 https://huggingface.co/papers/2609.24220 https://huggingface.co/papers/2609.25053 https://huggingface.co/papers/2609.25636 https://huggingface.co/papers/2609.25853 https://huggingface.co/papers/2609.26489 https://huggingface.co/papers/2609.26781 https://huggingface.co/papers/2609.29845 https://huggingface.co/papers/2609.33987 https://huggingface.co/papers/2609.36322 https://huggingface.co/papers/2610.00972 https://huggingface.co/papers/2610.01509 https://huggingface.co/papers/2610.01936 https://huggingface.co/papers/2610.10170 https://huggingface.co/papers/2610.11899 https://huggingface.co/papers/2610.12085 https://huggingface.co/papers/2610.12312 https://huggingface.co/papers/2610.12415 https://huggingface.co/zeroentropy/zembed-1-embedding

csdn (127) https://adg.csdn.net/695331d35b9f5f31781bc374.html https://adg.csdn.net/6a68c5bd662f9a54cb956060.html https://agent.csdn.net/6a52fd2f662f9a54cb8e8279.html https://blog.csdn.net/2301_79319949/article/details/160134772 https://blog.csdn.net/2301_80117363/article/details/163054080 https://blog.csdn.net/2401_84494441/article/details/160311466 https://blog.csdn.net/2401_85300269/article/details/164812712 https://blog.csdn.net/2401_87746054/article/details/162119032 https://blog.csdn.net/2403_87740391/article/details/160400735 https://blog.csdn.net/2501_92406411/article/details/158391326 https://blog.csdn.net/2611_95335967/article/details/164082693 https://blog.csdn.net/AgentInsight/article/details/162166516 https://blog.csdn.net/CYYYCDZ/article/details/163500980 https://blog.csdn.net/FuncWander/article/details/160794992 https://blog.csdn.net/Maolei_NyaRu_/article/details/163861971 https://blog.csdn.net/Python_cocola/article/details/161121337 https://blog.csdn.net/a28752405254060/article/details/156463704 https://blog.csdn.net/a2875254060/article/details/156463704 https://blog.csdn.net/a2875255254060/article/details/156463704 https://blog.csdn.net/chen1415886044/article/details/166736574 https://blog.csdn.net/enheng1238/article/details/163363550 https://blog.csdn.net/gedonshen/article/details/164287845 https://blog.csdn.net/gitblog_01200/article/details/155806102 https://blog.csdn.net/huangmingleiluo/article/details/149409536 https://blog.csdn.net/keshi_curry/article/details/160235298 https://blog.csdn.net/llwszx/article/details/163446336 https://blog.csdn.net/m0_59164520/article/details/160314591 https://blog.csdn.net/m0_59235245/article/details/162430133 https://blog.csdn.net/m0_59235945/article/details/162176846 https://blog.csdn.net/m0_69581581/article/details/162674291 https://blog.csdn.net/m0_69581581/article/details/162706987 https://blog.csdn.net/m0_69581581/article/details/163142070 https://blog.csdn.net/m0_69581581/article/details/163481543 https://blog.csdn.net/m0_69581581/article/details/163542988 https://blog.csdn.net/m0_69581581/article/details/163842030 https://blog.csdn.net/qcx23/article/details/160820786 https://blog.csdn.net/qcx23/article/details/160959849 https://blog.csdn.net/qq_27504375/article/details/163024278 https://blog.csdn.net/qq_31429225/article/details/163084256 https://blog.csdn.net/qq_33957603/article/details/165118957 https://blog.csdn.net/qq_35160742/article/details/163271979 https://blog.csdn.net/qq_40999403/article/details/148424108 https://blog.csdn.net/qq_45657541/article/details/163280072 https://blog.csdn.net/qq_56999332/article/details/161400644 https://blog.csdn.net/qq_73472828/article/details/161128801 https://blog.csdn.net/smallym1/article/details/163407419 https://blog.csdn.net/srk950606/article/details/156913534 https://blog.csdn.net/weixin_27310985/article/details/164650777 https://blog.csdn.net/weixin_29045585/article/details/163284405 https://blog.csdn.net/weixin_29046611/article/details/163992955 https://blog.csdn.net/weixin_31062533/article/details/166041904 https://blog.csdn.net/weixin_33212799/article/details/164766449 https://blog.csdn.net/weixin_33806859/article/details/165183806 https://blog.csdn.net/weixin_42509513/article/details/160697415 https://blog.csdn.net/weixin_42561143/article/details/160044073 https://blog.csdn.net/weixin_43705457/article/details/163801743 https://blog.csdn.net/weixin_43726381/article/details/161983901 https://blog.csdn.net/weixin_47547625/article/details/161535116 https://blog.csdn.net/weixin_52208686/article/details/162926745 https://blog.csdn.net/weixin_52208686/article/details/163067235 https://blog.csdn.net/weixin_53920044/article/details/148875486 https://blog.csdn.net/weixin_58753619/article/details/161024790 https://blog.csdn.net/xx_nm98/article/details/162153537 https://blog.csdn.net/xx_nm98/article/details/163927291 https://blog.csdn.net/yonggeit/article/details/163207114 https://blog.csdn.net/yonggeit/article/details/166773439 https://blog.csdn.net/zxsxj/article/details/161504705 https://codearts.csdn.net/6a92d6c63bda720d4b3799c5.html https://csdn.net/2301_77803292/article/details/162705936 https://csdn.net/article/2026-07-30/163342366 https://csdn.net/huang9537638381/article/details/157699556 https://csdn.net/kaka0722ww/article/details/161592550 https://csdn.net/likuolei/article/details/156766693 https://csdn.net/m0_59164520/article/details/161121236 https://csdn.net/weixin_33269743/article/details/162160811 https://csdn.net/weixin_35774598/article/details/162381479 https://csdn.net/weixin_37763484/article/details/148407621 https://csdn.net/weixin_42118737/article/details/149105217 https://csdn.net/yangshangwei/article/details/155892912 https://edu.csdn.net/course/detail/37529 https://gitcode.csdn.net/6a08f72f10ee7a33f2730f92.html https://gitcode.csdn.net/6a0eb9cb10ee7a33f274216e.html https://mcp.csdn.net/6a85893c662f9a54cb9e944e.html https://pixle.blog.csdn.net/article/details/161263511 https://www.csdn.net/yonggeit/article/details/162179765 https://xingyun3d.csdn.net/69f019380a2f6a37c5a67fba.html https://bbs.csdn.net/weixin_28717939/article/details/100541897 https://bbs.csdn.net/weixin_28745309/article/details/100561051 https://bbs.csdn.net/weixin_28840591/article/details/100538514 https://bbs.csdn.net/weixin_29047545/article/details/100557268 https://bbs.csdn.net/weixin_29055699/article/details/100563734 https://bbs.csdn.net/weixin_29062613/article/details/100548895 https://bbs.csdn.net/weixin_30985047/article/details/100536234 https://bbs.csdn.net/weixin_28221501/article/details/100567244 https://bbs.csdn.net/weixin_28839549/article/details/100569284 https://bbs.csdn.net/weixin_29038155/article/details/100486684 https://bbs.csdn.net/weixin_29040367/article/details/100507316 https://bbs.csdn.net/weixin_31715171/article/details/100488013 https://bbs.csdn.net/weixin_29031057/article/details/100536457 https://bbs.csdn.net/weixin_42691388/article/details/100476377 https://decodingaimagazine.substack.com/p/pre-inference-rag-evaluation-metrics https://itnews.substack.com/p/itne-cloud-devops-operations-security https://wrk3.substack.com/p/my-year-with-agents-and-why-i-refuse https://packtdatapro1.substack.com/p/more-ai-agents-more-problems-heres https://aws.amazon.com/blogs/machine-learning/building-trust-in-ai-agents-a-production-ready-evaluation-framework-with-amazon-bedrock-agentcore https://p4sc4l.substack.com/api/v1/file/f7964e41-8dac-4c42-a081-700273fffb8d.pdf https://bbs.csdn.net/weixin_28837817/article/details/100590113 https://blog.csdn.net/2301_79028807/article/details/168342939 https://blog.csdn.net/2301_79028807/article/details/168544552 https://agent.csdn.net/6a4e6d8a662f9a54cb8c0b63.html https://bbs.csdn.net/weixin_28221501/article/details/100625127 https://csdn.net/m0_69581581/article/details/163542988 https://csdn.net/m0_69581581/article/details/163842030 https://www.csdn.net/2301_77803292/article/details/162705936 https://www.csdn.net/a2875254060/article/details/156463704 https://www.csdn.net/huang9537638381/article/details/157699556 https://www.csdn.net/kaka0722ww/article/details/161592550 https://www.csdn.net/likuolei/article/details/156766693 https://www.csdn.net/m0_59164520/article/details/161121236 https://www.csdn.net/m0_69581581/article/details/163542988 https://www.csdn.net/weixin_33269743/article/details/162160811 https://www.csdn.net/weixin_35774598/article/details/162381479 https://www.csdn.net/weixin_37763484/article/details/148407621 https://www.csdn.net/weixin_42118737/article/details/149105217 https://www.csdn.net/yangshangwei/article/details/155892912 https://tianqi.csdn.net/6ad8b995536f457592b164b4.html https://blog.csdn.net/weixin_29055699/article/details/168575246

substack (238) https://abhijayvuyyuru.substack.com/p/the-free-8-week-roadmap-to-become-bea https://aiagentssimplified.substack.com/p/the-2026-path-to-learning-ai-agents https://aiamastery.substack.com/p/lesson-44-evaluating-agentic-rag https://aiengineeringinsider.substack.com/p/rag-interview-q-and-a-how-retrieval https://aiexpjourney.substack.com/p/advanced-rag-06-exploring-query-rewriting-23997297f2d1 https://aishwaryasrinivasan.substack.com/p/all-you-need-to-know-about-rag-in https://alphasignalai.substack.com https://alphasignalai.substack.com/p/rag-and-long-context-arent-enough https://bhavishyapandit9.substack.com https://businessanalytics.substack.com/p/freemium-what-the-latest-chain-of https://claudiostamile.substack.com/p/agent-memory-is-not-rag-a-practical https://cobusgreyling.substack.com/p/ragtruth https://designgurus.substack.com/p/llm-inference-at-scale-batching-caching https://dkennetz.substack.com/p/llm-inference-curriculum https://engineersguide.substack.com/p/best-vector-databases-rag https://gradientflow.substack.com/p/rag-reimagined-5-breakthroughs-you https://inferenceengineer.substack.com/p/understanding-graph-data-mcps https://jamwithai.substack.com/p/the-2026-roadmap-production-aiml https://kenhuangus.substack.com/p/announcing-the-10-part-series-the https://kenhuangus.substack.com/p/how-ai-agents-actually-remember-inside https://kenhuangus.substack.com/p/the-kv-cache-wars https://micheallanham.substack.com/p/comparative-analysis-of-rag-architectures https://open.substack.com/pub/alexewerlof/p/owasp-top-10-ai-llm-agents https://open.substack.com/pub/alexewerlof/p/owasp-top-10-ai-llm-agents?r=18prwo https://open.substack.com/pub/pragmaticengineer/p/what-is-inference-engineering https://pawankjha.substack.com/p/architecting-llm-inference-part-1 https://priyankavergadia.substack.com/p/10-types-of-rag-you-need-to-know https://substack.com/@systemdesignone/note/c-359913029 https://theaiengineer.substack.com/p/the-ai-agents-stack-2026-edition https://theaiengineer.substack.com/p/what-is-agent-memory https://theaiengineer.substack.com/p/what-is-rag-retrieval-augmented-generation https://theaiengineer.substack.com/p/why-ai-agents-keep-failing-in-production https://businessanalytics.substack.com/p/freemium-what-the-latest-chain-of https://heathercoxrichardson.substack.com/p/october-14-2026/comments https://davidorban.substack.com/p/is-extraction-economics-coming-to other(190) https://1bench.dev/databases/vector https://academy.dair.ai/papers/does-your-agents-memory-survive-a-model-upgrade-a-controlled-study-of-memory-por-2609.05339 https://aclanthology.org/2025.findings-acl.861 https://aclanthology.org/2026.findings-acl.1619.pdf https://actian.com/blog/databases/how-to-evaluate-vector-databases-in-2026 https://agentmarketcap.ai/blog/2026/04/11/agent-memory-architecture-production-2026 https://aiquinta.ai/blog/harness-engineering-guide-to-reliable-ai-agents https://alphacorp.ai/blog/best-vector-databases-for-rag-2026-top-7-picks https://aminrj.com/posts/rag-security-architecture https://anthropic.com/engineering/effective-context-engineering-for-ai-agents https://appinventiv.com/blog/how-to-build-custom-ai-agents https://arize.com/ai-agents https://atlan.com/know/advanced-rag-techniques https://atlan.com/know/best-ai-agent-memory-frameworks-2026 https://atlan.com/know/what-is-rag https://blog.bytebytego.com/p/graphrag-how-ai-answers-questions https://blog.compendialabs.org/posts/2026-09-07-al-agent-memory-model-upgrades https://blog.starmorph.com/blog/rag-techniques-compared-best-practices-guide https://blog.turingpost.com/p/ragtypes https://buzzgrewal.medium.com/ai-agents-dont-need-vector-search-anymore-inside-the-agentic-search-stack-replacing-rag-in-2026-58efcabe4f6f https://c3.unu.edu/projects/ai/deepresearch/demo_research-report_searchengine2.html https://ceur-ws.org/Vol-4211/paper05.pdf https://codefarm.in/blog/gen-ai/long-context-vs-rag-2026 https://commandcode.ai/guides/rag-in-2026 https://cometragobservability.com/2026-tools https://cruxdigits.nl/blog/rag-tables-bm25-vs-vector-search https://cruxdigits.nl/blog/rag-vs-graphrag-2026 https://www.designveloper.com/blog/best-open-source-vector-database https://www.cake.ai/blog/why-90-of-agentic-rag-projects-fail-and-how-cake-changes-that https://cycode.com/blog/owasp-top-10-agentic-applications https://datanucleus.dev/rag-and-agentic-ai/agentic-rag-enterprise-guide-2026 https://dataiku.com/learn/rag/rag-evaluation-production https://dev.to/gabrielanhaia/70-of-enterprise-rag-deployments-fail-before-production-heres-what-kills-them-26ml https://dev.to/varun_pratapbhardwaj_b13/5-ai-agent-memory-systems-compared-mem0-zep-letta-supermemory-superlocalmemory-2026-benchmark-59p3 https://developer.jdcloud.com/article/4382 https://developers.redhat.com/articles/2026/04/23/deploy-openviking-openshift-ai-improve-ai-agent-memory https://docs.oracle.com/en-us/iaas/releasenotes/services/conteng https://doi.org/10.48550/arxiv.2607.26760 https://duynguyenngoc.com/posts/advanced-rag-techniques-2026 https://easychair.org/cfp/SynIRgy2026 https://evoailabs.medium.com/metis-the-dawn-of-native-memory-in-foundation-models-de4ddb55cef1 https://futureagi.com/blog/agentic-rag-systems-2025 https://futureagi.com/blog/llm-agent-architectures-core-components/ https://futureagi.com/blog/rag-architecture-llm-2025 https://futureagi.com/blog/rag-summarization https://genai.owasp.org/resource/owasp-top-10-for-agentic-applications-for-2026 https://getfeather.store/theory/agentic-search-vs-vector-database-2026 https://goteleport.com/blog/owasp-top-10-agentic-applications https://gradientflow.com/eight-security-assumptions-ai-agents-quietly-break/ https://gradientflow.com/six-reasons-i-think-open-ai-models-will-win/ https://gradientflow.com/the-ai-data-problem-moved-downstream/ https://gradientflow.com/the-part-of-the-ai-security-story-i-keep-thinking-about/ https://gradientflow.com/your-model-is-easy-to-buy-your-data-pipeline-is-not/ https://harness.io/blog/ai-deployment-in-production-orchestrate-llms-rag-agents https://hyper.ai/en/papers/2608.24053 https://iclr.cc/virtual/2026/oral/10010224 https://investors.progress.com/news-releases/news-release-details/progress-agentic-rag-named-2026-ai-excellence-award-winner https://jobsbyculture.com/jobs/ai-engineer-memory-systems https://kanerika.com/blogs/rag-vs-agentic-rag https://lilianweng.github.io/posts/2026-07-04-harness/ https://llm-d.ai https://llms3.com/blog/when-the-benchmarks-stopped-agreeing-july-2026 https://local-ai-zone.github.io/guides/best-ai-reranker-models-ultimate-ranking-2026.html https://lyzr.ai/blog/agentic-rag https://mc-search-project.github.io https://medium.com/@Micheal-Lanham/pipeline-rag-vs-agentic-rag-vs-knowledge-graph-rag-what-actually-works-and-when-47a26649a457 https://medium.com/coding-nexus/wemm-embedding-the-open-source-multimodal-embedding-model-built-for-1-billion-requests-a-day-e8cd1cbb518f https://mem0.ai/blog/ai-memory-benchmarks-in-2026 https://mem0.ai/blog/state-of-ai-agent-memory-2026 https://mem0.ai/research https://memoryarena.github.io https://microsoft.com/en-us/research/blog/echoverse-deep-evolving-environments-for-computer-use-agents/ https://microsoft.com/en-us/research/blog/evolib-turning-experience-into-evolving-knowledge/ https://microsoft.com/en-us/research/blog/orchard-an-open-framework-for-scalable-agentic-ai/ https://mlcommons.org/cluster/2026/08/endtoend-inference https://mlflow.org/articles/llm-application-architecture-a-2026-engineers-guide/ https://mnemoverse.com/docs/library/ai-memory-solutions-2026-q3 https://multimodalrag.github.io https://neo4j.com/videos/nodes-2026-memmachine-agents-that-learn-memory-that-lasts https://neo4j.com/videos/nodes-2026-the--agent-memory-landscape https://neo4j.com/videos/nodes-2026-the-agent-memory-landscape https://neo4j.com/videos/nodes-2026-the-ai-agent-memory-landscape https://noqta.tn/en/blog/harness-engineering-ai-agents-production-ready-2026 https://ojs.aaai.org/index.php/AAAI/article/view/40726/44687 https://onyx.app/insights/enterprise-rag-platforms-2026 https://openreview.net/forum?id=GROUNDEDKG-RAG https://openreview.net/forum?id=ParliamentRAG https://preuve.ai/blog/ai-memory-systems-statistics-2026 https://proceedings.iclr.cc/paper_files/paper/2026/file/fd1eff9dd295df50a41f2521942fa31d-Paper-Conference.pdf https://promptfoo.dev/blog/rag-poisoning https://pub.towardsai.net/all-you-need-to-know-about-chunking-in-agentic-rag-bdc34c21a6e6 https://pub.towardsai.net/mastering-agentic-rag-3-architecture-patterns-for-production-grade-ai-system-with-examples-03f799a3cbd0 https://pub.towardsai.net/the-agent-context-wars-three-battles-at-different-layers-487dab110ea8 https://ragaboutit.com https://ragaboutit.com/7-rag-failure-modes-crippling-enterprise-deployments-in-2026 https://ranksquire.com/2026/02/26/best-vector-database-rag-applications-2026 https://rauljitechnologies.com/blog/agentic-rag-2026 https://research.google/blog/chain-of-agents-large-language-models-collaborating-on-long-context-tasks https://research.google/blog/unlocking-dependable-responses-with-gemini-enterprise-agent-platforms-agentic-rag/ https://semanticscholar.org/paper/SoK-Agentic-RAG-Mishra-Niroula/e23f86a3021bb728af4a2d7f70cccb50ff91f4eb https://shattered.io/hugging-face-agent-intrusion-anatomy-2026 https://simo-online.com/en/blog/rag-retrieval-augmented-generation-unternehmen-leitfaden-2026 https://skywork.ai/slide/en/harnessing-ai-agents-2056955859996663809 https://slash-digital.io/en/insights/rag-2026 https://spheron.network/blog/vllm-vs-sglang-2026 https://splx.ai/blog/rag-poisoning-in-enterprise-knowledge-sources https://squirro.com/squirro-blog/state-of-rag-genai https://supermemory.ai/blog/hybrid-search-guide https://tech.yahoo.com/ai/article/openai-just-disclosed-more-concerning-ai-behavior-following-the-hugging-face-incident-heres-everything-you-need-to-know-154529895.html https://thenewstack.io/graphrag-multi-hop-reasoning-python https://tianpan.co/guides/rag-vs-long-context-decision-framework https://turingpost.com/p/ragtypes https://usewire.io https://varunpratap.com/blog/superlocalmemory-vs-mem0-benchmark-2026 https://vectorize.io/articles/owasp-asi06 https://wavect.io/blog/rag-vs-finetune-vs-longcontext-2026 https://workativ.com/hr/blog/agentic-rag https://www.actian.com/blog/databases/how-to-evaluate-vector-databases-in-2026 https://www.alphaxiv.org/abs/2604.04936 https://www.alphaxiv.org/abs/2604.09666 https://www.alphaxiv.org/abs/2607.26497 https://www.amazon.science/publications/keyword-search-is-all-you-need-achieving-rag-level-performance-without-vector-databases-using-agentic-tool-use https://www.articsledge.com/post/lazygraphrag-retrieval-augmented-generation https://www.cake.ai/blog/why-90-of-agentic-rag-projects-fail-and-how-cake-chives-that https://www.codecon.sk/en/whats-new-in-rag https://www.databricks.com/blog/long-context-rag-performance-llms https://www.designvelop.com/blog/best-open-source-vector-database https://www.digitalapplied.com/blog/88-percent-ai-agents-never-reach-production-failure-framework https://www.digitalapplied.com/blog/hybrid-search-bm25-vector-reranking-reference-2026 https://www.elastic.co/search-labs/blog/jina-reranker-35-legal-medical-structured-data https://www.emergentmind.com/topics/jina-reranker-v3 https://www.falkordb.com/blog/how-graph-architecture-grounds-rag-and-prevents-hallucinations https://www.firecrawl.dev/blog/best-chunking-strategies-rag https://www.getfeather.store/theory/agentic-search-vs-vector-database-2026 https://www.glean.com/perspectives/top-10-enterprise-use-cases-for-rag-models-in-2026 https://www.harness.io/blog/ai-deployment-in-production-orchestrate-llms-rag-agents https://www.lancedb.com/blog/hybrid-search-rag-for-real-life-production-grade-applications-e1e727b3965a https://www.linkedin.com/posts/muhammad-arslan-b591b1285_research-aiagents-agenticai-activity-7474154906019438592-erLD https://www.lyzr.ai/blog/agentic-rag https://www.meilisearch.com/blog/graph-rag https://www.microsoft.com/en-us/research/blog/lazygraphrag-setting-a-new-standard-for-quality-and-cost https://www.open-techstack.com/blog/pgvector-vs-qdrant-2026 https://www.paperclipped.de/en/blog/graph-rag-production https://www.percona.com/blog/vector-distance-vector-similarity-search-in-percona-server-for-mysql https://www.sphereinc.com/blogs/hybrid-search-enterprise-rag https://www.spheron.network/blog/vllm-vs-sglang-2026 https://www.strike48.com/post/agentic-ai-security-risks https://www.turingpost.com/p/ragtypes https://www.varunpratap.com/blog/superlocalmemory-vs-mem0-benchmark-2026 https://www.youtube.com/watch?v=9BpXVodhALk https://www.youtube.com/watch?v=ZNFFuzuwSBc https://www.youtube.com/watch?v=rEITYxTJggU https://www.zartis.com/rag-in-production-what-nobody-tells-you-until-youre-debugging-it https://x.com/hwchase17/status/2105710409894547687 https://x.com/jerryjliu0/status/2105692426577056106 https://x.com/liquidai/status/2105003472332693869 https://x.com/maximelabonne https://x.com/omarsar0/status/2106423166788669445 https://x.com/omarsar0/status/2106927371366596692 https://x.com/omarso/2106700905051746803 https://x.com/rasbt/status/2106369670680871031 https://huxai.substack.com/p/from-research-to-evidence https://github.com/intel/enterprise-agent-toolkit https://github.com/topics/ai-agent-framework https://github.com/vstorm-co/agenticos https://recsys.substack.com/i/220663591/1-jevgpt-thinking-fast-and-slow-in-autonomous-research https://recsys.substack.com/i/220663591/6-personalized-llm-recommendation-with-semantic-ids-autoregressive-beam-search-vs-multi-token-prediction https://recsys.substack.com/i/220663591/8-configuring-rag-systems-what-breaks-where-originates-and-how-to-fix https://neo4j.com/videos/nodes-ai-2026-memmachine-agents-that-learn-memory-that-lasts https://neo4j.com/videos/nodes-ai-2026-the-agent-memory-landscape https://neo4j.com/videos/nodes-ai-2026-the-ai-agent-memory-landscape https://open-techstack.com/blog/pgvector-vs-qdrant-2026 https://www.krishnaik.in/project/real-time-voice-ai-agent-with-rag-and-low-latency-voice-processing https://www.parloa.com/knowledge-hub/what-is-rag-in-ai https://blog.bytebytego.com/p/ep229-8-chunking-strategies-for-rag https://blog.bytebytego.com/feed https://aws.amazon.com/blogs/machine-learning/introducing-reflection-beam-an-open-weight-model-for-agentic-coding-on-aws https://doi.org/10.48550/arxiv.2606.14061 https://doi.org/10.48550/arxiv.2510.09665 https://doi.org/10.48550/arxiv.2511.02230 https://www.orcarouter.ai/blog/step-5-preview-serving-leak https://saiyampathak.substack.com/p/30000-merges-a-day-anthropic-left https://gradientflow.substack.com/p/are-you-building-a-model-or-a-model https://consultingedge.substack.com/api/v1/file/dadff746-10cf-41ea-a71d-e0379d78871c.pdf https://blog.bytebytego.com/p/ep230-rag-vs-agentic-rag https://aclanthology.org/2026.findings-acl.78 https://aclanthology.org/2026.findings-acl.78.pdf https://www.trydeepteam.com/guides/guide-red-teaming-agentic-rag https://www.redfoxsec.com/blog/rag-pipeline-attack-scenarios-context-manipulation-and-data-leakage-explained https://developer.ibm.com/articles/agentic-rag-pipeline https://swyx.substack.com/p/claude-mods https://javarevisited.substack.com/p/30-essential-ai-engineering-concepts https://machinelearningwithme.com/agent/chapter05/short-term-memory-rag-agent/ https://aclanthology.org/2026.findings-acl.648 https://aclanthology.org/2026.findings-acl.648.pdf https://doi.org/10.18653/v1/2026.findings-acl.648 https://www.ovaledge.com/blog/agentic-rag-governance https://www.taskade.com/blog/llm-rag-agent-agentic https://www.relevancelab.com/ai-services/rag-development https://www.cekura.ai/discover/datasets-training-evaluating-rag-models https://smile.eu/en/node/878 https://onepixelsoft.com/blog-details/rag-in-2026-evolution-future-of-ai-retrieval

本次变更

R129 第 129 轮（2026-10-26 13:55 CST）：0 件 RAG 主分类 arXiv 净增 + 8 件增量材料（A 组 5 repo_cards takeshy/obsidian-local-llm-hub 90⭐ 本地 Obsidian RAG hub + api-evangelist/pienso 0⭐ + Ablation-Tool/ablation 223⭐ AIARE 框架 + devrochy/agent-core 0⭐ per-tenant AgentSpec + Postgres RLS + crmne/ruby_llm 4483⭐ +11 周增 Ruby 框架 + B 组 Jay 技术博客与开源动态 6 条（已核 200：IBM agentic RAG pipeline CrewAI 四智能体 + 意图分类 + 混合数据层 / swyx Claude Mods；搜索线索待核 4 条：NirDiamant/RAG_Techniques + PacktPublishing/Building-Production-Ready-AI-Agents + microsoft/ai-agent-runbooks + Java Revisited 30 AI Engineering Concepts）+ C 组 Jay CSDN 2 条（已核 200：短期记忆 RAG Agent 删除/修剪/总结三策略 + StateGraph + PostgresSaver checkpointer；未核 1 条 521 未重试）+ D 组 1 paper_card 2001.08361 Scaling Laws 分类体系错位命中）+ 1 件新 arXiv ID + 2 件新 DOI（10.48550/arxiv.2001.08361 + 10.18653/v1/2026.findings-acl.648）+ 0 新 CVE + web_search 补 2 件 fresh（Reflective RAG ACL Findings 2026 SFT 内部不确定性自校正 + RL 动态 KL 正则；Taskade 引 Infragistics 2026-07 RAG vs agentic retrieval 100 查询成本/延迟/胜率/引文对照）+ 67 维张力续立 + 共识 236 扩写/争议 198 续立并再次重述 + 开放 279 补自校正标签方法论参照/281 补首 token 延迟成本证据 + 284 实质推进（收窄为委托链上的授权主体传递）/285 部分推进（verify 判据候选答案）+ 趋势 343 扩写为可核查性进入可训练时代（四维齐涨）。

Tom · 2026-10-26 13:55 CST · R129 Wave3 E1 · 0 件主分类净增 + 8 件增量 + 1 件新 arXiv ID + 2 件新 DOI + 共识 236/争议 198/开放 279/281/284/285/趋势 343 + 67 维张力㊿̲+0

R128 第 128 轮（2026-10-25 13:55 CST）：1 件 RAG 主分类 arXiv 净增 + 8 件增量材料（A 组 3 repo_cards academic-writing 簇 parthchandak02/literature-review-assistant 1⭐ PRISMA 2020 系统综述全流程 + YenochQin/llm-wiki 1⭐ Zotero+MinerU+知识图谱写作 + Aryan-Pardeshi/CiteProof-AI-MCP 89⭐ +7 周增 24 个 MCP 审计工具 + B 组 Jay ByteByteGo RSS 1 条 EP230 RAG vs Agentic RAG + C 组 Tom 文献雷达 1 篇 2610.30418 Agent Mosaic + D 组 3 paper_cards 2301.12597 BLIP-2 + 1506.06726 Skip-Thought Vectors + 2007.14062 Big Bird 分类体系错位命中）+ 4 件新 arXiv ID + 4 件新 DOI + 0 新 CVE + web_search 补 1 件 fresh ACL Findings 2026.findings-acl.78 Agentic RAG 数据中心综述 + 67 维张力续立 + 共识 236 扩写/争议 198 续立并放宽一格 + 开放 279 补禁止集方法论参照/281 补待查线索 + 新增 284 RAG 安全分析单位之争（逐文档 vs 复合结论级）+ 新增 285 Agentic RAG 数据构造开放 + 趋势 343 重写为可核查性分解为权限与引证两条线。

Tom · 2026-10-25 13:55 CST · R128 Wave3 E1 · 1 件主分类净增 + 8 件增量 + 4 件新 arXiv ID + 4 件新 DOI + 共识 236/争议 198/开放 279/281/284/285/趋势 343 + 67 维张力㊿̲+0

R127 第 127 轮（2026-10-22 13:55 CST）：0 件 RAG 主分类 arXiv 净增 + 9 件增量材料（A 组 3 repo_cards 首次采集 HyxiaoGe/fusion 102⭐ experimental + worldbench/awesome-ai-auto-research 561⭐ research 待 LLM 分类 + iblai/os 821⭐ research 企业 Agent 平台 + B 组 Jay cs.IR RSS 5 条 2610.28569 ECHO + 2610.28435 多模态重排系统决策 + 2610.28054 AutoSim + 2610.27898 SPAR 前缀平衡 SID + 2610.27850 Pareto 联合评估鲁棒性 + C 组 Jay 工程简报保留 2 丢弃 0 + D 组 Jay 工程实践筛选保留 2 丢弃 4 + E 组 Jay 博客发现 1 线索 + F 组 Jay CSDN 2 已核 6 待查）+ 5 件新 arXiv ID + 0 件新 DOI + 0 新 CVE + 67 维张力续立 + 共识 236 扩写/争议 198 续立/开放 279 增量 + 281 补工程侧成本证据 + 新增 282 SID 稳定性 vs 前缀平衡相容性 + 新增 283 Airbnb 以可移植 Agent Skill 替换生产 RAG 栈实证缺口 + 趋势 343 重写为评测批判成建制。

Tom · 2026-10-22 13:55 CST · R127 Wave3 E1 · 0 件主分类净增 + 9 件增量 + 5 件新 arXiv ID + 共识 236/争议 198/开放 279/281/282/283/趋势 343 + 67 维张力㊿̲+0