<![CDATA[Lynn121的博客]]>

<![CDATA[Lynn121的博客]]>https://blog.youkuaiyun.com/Lynn121zh-cnhttps://blog.youkuaiyun.com/<![CDATA[仿真到现实(Sim-to-Real)迁移:跨越仿真与现实鸿沟一、仿真训练的价值与现实落差二、域随机化:以多样性覆盖不确定性三、域自适应与系统辨识四、真实数据回填与混合训练五、评估、风险与]]>https://blog.youkuaiyun.com/Lynn121/article/details/170246692https://blog.youkuaiyun.com/Lynn121/article/details/170246692Lynn121Sat, 07 Nov 2026 20:14:44 +0800<![CDATA[触觉与多模态感知:具身智能的感官基础一、为什么具身智能离不开触觉二、触觉感知的技术路径三、多模态融合的必要性四、感知一行动闭环中的学习五、通向灵巧操作的感知挑战#多模态#具身智能#人工智]]>https://blog.youkuaiyun.com/Lynn121/article/details/170201164https://blog.youkuaiyun.com/Lynn121/article/details/170201164Lynn121Fri, 06 Nov 2026 22:44:15 +0800<![CDATA[生成式世界模型与视频预测:让机器想象未来一、世界模型为何需要“生成“能力二、视频预测作为世界模型的训练场三、潜空间表征与可控生成四、世界模型如何服务决策五、局限与通向更可靠的世界模型#世界]]>https://blog.youkuaiyun.com/Lynn121/article/details/170148483https://blog.youkuaiyun.com/Lynn121/article/details/170148483Lynn121Thu, 05 Nov 2026 20:46:31 +0800<![CDATA[自然语言处理的技术演进与核心任务一、技术演进的主线二、语言理解类任务三、语言生成类任务四、跨语言与多模态五、应用边界与负责任使用#自然语言#自然语言处理#人工智能#大模型]]>https://blog.youkuaiyun.com/Lynn121/article/details/170051240https://blog.youkuaiyun.com/Lynn121/article/details/170051240Lynn121Wed, 04 Nov 2026 18:50:27 +0800<![CDATA[语音识别与语音合成的技术原理及应用一、语音识别的核心原理二、语音合成的核心原理三、工程落地的主要难点四、典型应用场景五、需要关注的现实问题#语音识别#语音合成#人工智能#大模型]]>https://blog.youkuaiyun.com/Lynn121/article/details/169951730https://blog.youkuaiyun.com/Lynn121/article/details/169951730Lynn121Tue, 03 Nov 2026 21:43:45 +0800<![CDATA[知识工程的构建方法与落地最佳实践一、知识工程的核心理念二、知识获取与清洗三、知识组织与建模四、质量控制与治理五、持续运营与演进#知识工程#人工智能#大模型#世界模型]]>https://blog.youkuaiyun.com/Lynn121/article/details/169650996https://blog.youkuaiyun.com/Lynn121/article/details/169650996Lynn121Sat, 31 Oct 2026 18:47:17 +0800<![CDATA[大模型轻量化的核心路径与技术选型1/1 轻量化的动机与总览二、知识蒸馏三、量化与低比特推理四、剪枝与结构精简五、低秩分解与组合策略#大模型轻量化#人工智能#大模型#知识蒸馏]]>https://blog.youkuaiyun.com/Lynn121/article/details/169564317https://blog.youkuaiyun.com/Lynn121/article/details/169564317Lynn121Fri, 30 Oct 2026 22:20:25 +0800<![CDATA[大模型检索增强生成(RAG)的技术原理与架构设计一、检索增强生成的基本思想二、RAG系统的核心架构三、检索质量的关键影响因素四、知识融合与可信生成五、工程落地中的常见挑战#大模型#RAG#]]>https://blog.youkuaiyun.com/Lynn121/article/details/169462846https://blog.youkuaiyun.com/Lynn121/article/details/169462846Lynn121Thu, 29 Oct 2026 18:47:54 +0800<![CDATA[知识图谱:结构化表达实体与关系的知识组织技术一、知识图谱基础概念与数据结构二、知识图谱构建的核心流程三、知识推理:挖掘图谱中隐含关联四、知识图谱与大模型的能力互补性五、图谱与大模型融合应用趋]]>https://blog.youkuaiyun.com/Lynn121/article/details/169367161https://blog.youkuaiyun.com/Lynn121/article/details/169367161Lynn121Wed, 28 Oct 2026 21:54:24 +0800<![CDATA[具身智能与边缘AI的相遇:机器人如何“轻装上路“一、机器人需要“就地思考“二、轻量化为机器人“减负三、端云协同的机器人方案四、机器人应用的前沿探索五、工程师的“新工具箱#具身智能#人工智能]]>https://blog.youkuaiyun.com/Lynn121/article/details/169269236https://blog.youkuaiyun.com/Lynn121/article/details/169269236Lynn121Tue, 27 Oct 2026 23:07:54 +0800<![CDATA[人工智能大模型应用工程师:连接技术与业务的新桥梁一、为什么需要这样一位“翻译官“二、岗位的核心能力清单三、从想法到落地的完整旅程四、不同行业的应用实践五、成长路径与未来空间#人工智能]]>https://blog.youkuaiyun.com/Lynn121/article/details/169168986https://blog.youkuaiyun.com/Lynn121/article/details/169168986Lynn121Mon, 26 Oct 2026 20:46:46 +0800<![CDATA[大模型检索增强与知识工程:让大模型“言之有据“一、大模型为何需要“外挂记忆“二、检索增强的核心流程三、知识工程的角色四、落地场景与价值五、把“知识“变成竞争力#大模型#知识工程#人工智能]]>https://blog.youkuaiyun.com/Lynn121/article/details/168979359https://blog.youkuaiyun.com/Lynn121/article/details/168979359Lynn121Sat, 24 Oct 2026 23:59:04 +0800<![CDATA[具身智能:让人工智能拥有物理世界感知能力一、具身智能的核心思想与理念起源二、具身智能系统的三大组成环节三、具身智能的典型应用方向四、具身智能发展的主要难点五、具身智能的长期演进趋势#具身智]]>https://blog.youkuaiyun.com/Lynn121/article/details/168884756https://blog.youkuaiyun.com/Lynn121/article/details/168884756Lynn121Fri, 23 Oct 2026 23:33:47 +0800<![CDATA[数字孪生走进日常生活:从城市治理到个人健康一、城市:在虚拟中预演现实二、家庭:让居住空间“善解人意“三、健康:从被动就医到主动管理四、生活场景背后的技术支撑五、让技术真正为人服务#数字孪生]]>https://blog.youkuaiyun.com/Lynn121/article/details/168662388https://blog.youkuaiyun.com/Lynn121/article/details/168662388Lynn121Wed, 21 Oct 2026 21:45:45 +0800<![CDATA[模型轻量化技术:压缩、蒸馏与低秩分解的实现路径一、模型轻量化的背景与价值二、模型压缩:删减冗余，精简体量三、知识蒸馏:迁移能力，打造小而强四、低秩分解:拆解权重矩阵，简化结构五、轻量化技术组]]>https://blog.youkuaiyun.com/Lynn121/article/details/168561647https://blog.youkuaiyun.com/Lynn121/article/details/168561647Lynn121Tue, 20 Oct 2026 22:59:03 +0800<![CDATA[【无标题】]]>https://blog.youkuaiyun.com/Lynn121/article/details/168454291https://blog.youkuaiyun.com/Lynn121/article/details/168454291Lynn121Mon, 19 Oct 2026 22:30:12 +0800<![CDATA[具身智能环境交互与学习范式原理一、具身智能交互的本质:行动带来环境状态变化二、强化学习范式:试错学习的底层原理三、模仿学习范式:从演示行为提取策略原理四、虚实迁移学习:模拟器预训练到真机微调]]>https://blog.youkuaiyun.com/Lynn121/article/details/168248766https://blog.youkuaiyun.com/Lynn121/article/details/168248766Lynn121Sat, 17 Oct 2026 20:27:31 +0800<![CDATA[【无标题】]]>https://blog.youkuaiyun.com/Lynn121/article/details/168143954https://blog.youkuaiyun.com/Lynn121/article/details/168143954Lynn121Fri, 16 Oct 2026 20:39:30 +0800<![CDATA[AI赋能产业升级与中美未来经济赛道博弈]]>https://blog.youkuaiyun.com/Lynn121/article/details/167703707https://blog.youkuaiyun.com/Lynn121/article/details/167703707Lynn121Mon, 12 Oct 2026 22:39:50 +0800<![CDATA[世界模型，是“空中楼阁”还是AGI的“最后一块拼图”？]]>https://blog.youkuaiyun.com/Lynn121/article/details/163252751https://blog.youkuaiyun.com/Lynn121/article/details/163252751Lynn121Mon, 27 Jul 2026 23:45:42 +0800