corpus.blog
41554 blogs · [ { "id": "01a0bb6b-3f88-700c-a512-4d230ca119be", "title": "新史记：CC外泄记", "url": "https://liweinlp.com/13514", "published_at": "2026-04-04T16:49:12+00:00" }, { "id": "01a0bb6b-3f88-700c-a512-4d230ca93a3f", "title": "关于Claude Code 泄漏事件的愚人节思考", "url": "https://liweinlp.com/13513", "published_at": "2026-04-02T20:58:21+00:00" }, { "id": "01a0bb6b-3f88-700c-a512-4d230d72e380", "title": "What the Claude Code TypeScript Leak Really Revealed", "url": "https://liweinlp.com/13512", "published_at": "2026-04-02T20:48:28+00:00" }, { "id": "01a0bb6b-3f88-700c-a512-4d230dab2a4d", "title": "When Agents Become the Default Gateway, Will the Operating System Be Rewritten?", "url": "https://liweinlp.com/13504", "published_at": "2026-02-13T06:17:36+00:00" }, { "id": "01a0bb6b-3f88-700c-a512-4d230dbe6fc7", "title": "When Agents Become the Default Gateway, Does the App Store Model Collapse?", "url": "https://liweinlp.com/13502", "published_at": "2026-02-13T05:50:43+00:00" }, { "id": "01a0bb6b-3f88-700c-a512-4d230eb27146", "title": "The Great Software Shake-Up of the Agent Era — Starting with OpenClaw", "url": "https://liweinlp.com/13506", "published_at": "2026-02-13T05:36:28+00:00" }, { "id": "01a0bb6b-3f88-700c-a512-4d230f4d18f8", "title": "Some Basic Agentic AI terminology", "url": "https://liweinlp.com/13499", "published_at": "2026-02-13T00:07:09+00:00" }, { "id": "01a0bb6b-3f88-700c-a512-4d230fade39e", "title": "OpenClaw as a case study of the coming Agentic AI era", "url": "https://liweinlp.com/13497", "published_at": "2026-02-12T15:49:03+00:00" }, { "id": "01a0bb6b-3f88-700c-a512-4d230fd8b0f4", "title": "Agent 时代的临界点：谈谈 OpenClaw 的安全隐患", "url": "https://liweinlp.com/13493", "published_at": "2026-02-12T15:25:49+00:00" }, { "id": "01a0bb6b-3f88-700c-a512-4d231051550c", "title": "Agent 时代的一些术语澄清", "url": "https://liweinlp.com/13490", "published_at": "2026-02-12T04:08:32+00:00" }, { "id": "01a0bb6b-3f88-700c-a512-4d2310c0001b", "title": "当 Agent 成为默认入口，App Store 模式是否崩塌？", "url": "https://liweinlp.com/13488", "published_at": "2026-02-12T03:59:36+00:00" }, { "id": "01a0bb6b-3f88-700c-a512-4d2311497d9c", "title": "当 Agent 成为默认入口，操作系统会不会被重写？", "url": "https://liweinlp.com/13487", "published_at": "2026-02-12T03:25:33+00:00" }, { "id": "01a0bb6b-3f88-700c-a512-4d2311acf4d3", "title": "Agent 时代的软件产业大洗牌：从 OpenClaw谈起", "url": "https://liweinlp.com/13486", "published_at": "2026-02-12T02:26:33+00:00" }, { "id": "01a0bb6b-3f88-700c-a512-4d2311e825da", "title": "2026年智能体范式大爆发：从认知幻象到工业化协同", "url": "https://liweinlp.com/13484", "published_at": "2026-02-11T08:32:12+00:00" }, { "id": "01a0bb6b-3f88-700c-a512-4d23126381ff", "title": "腾讯科技春节访谈，Agent 这一年：沸沸扬扬之后", "url": "https://liweinlp.com/13482", "published_at": "2026-02-10T08:42:55+00:00" }, { "id": "01a0bb6b-3f88-700c-a512-4d23131e9804", "title": "多模态大模型输入信号的离散化", "url": "https://liweinlp.com/13477", "published_at": "2026-02-01T02:46:57+00:00" }, { "id": "01a0bb6b-3f88-700c-a512-4d23134a793c", "title": "置顶:《立委关于大模型与AI的科学网博客汇总》", "url": "https://liweinlp.com/13476", "published_at": "2026-01-29T19:56:11+00:00" }, { "id": "01a0bb6b-3f88-700c-a512-4d2313904fbd", "title": "The Removal of Autopilot: A Misjudgment of Trust, Pricing Power, and Timing", "url": "https://liweinlp.com/13474", "published_at": "2026-01-28T04:03:46+00:00" }, { "id": "01a0bb6b-3f88-700c-a512-4d2313ad9996", "title": "Autopilot 被剥离: 一次关于信任与定价权的误判", "url": "https://liweinlp.com/13471", "published_at": "2026-01-28T02:28:45+00:00" }, { "id": "01a0bb6b-3f88-700c-a512-4d2314429b3e", "title": "自动驾驶已经解决了，但我们还没准备好告别驾驶", "url": "https://liweinlp.com/13470", "published_at": "2026-01-28T02:04:46+00:00" }, { "id": "01a0bb6f-2ce9-70f0-b717-5577bbb00d48", "title": "How FSD Quietly Took Control of Pricing Power", "url": "https://liweinlp.com/13469", "published_at": "2026-01-26T11:30:22+00:00" }, { "id": "01a0bb6f-2ce9-70f0-b717-5577bbc3de43", "title": "If Robotaxi Fails, This Is Where It Will Fail", "url": "https://liweinlp.com/13467", "published_at": "2026-01-26T11:19:37+00:00" }, { "id": "01a0bb6f-2ce9-70f0-b717-5577bc34d615", "title": "FSD 会拯救“最不被保险欢迎的人”", "url": "https://liweinlp.com/13466", "published_at": "2026-01-26T07:03:50+00:00" }, { "id": "01a0bb6f-2ce9-70f0-b717-5577bc72676a", "title": "保险降价，是自动驾驶第一次“自证盈利模型”", "url": "https://liweinlp.com/13465", "published_at": "2026-01-26T06:45:54+00:00" }, { "id": "01a0bb6f-2ce9-70f0-b717-5577bd160d03", "title": "如果 FSD 真的会失败，特斯拉最可能栽在哪里？", "url": "https://liweinlp.com/13464", "published_at": "2026-01-25T21:55:45+00:00" }, { "id": "01a0bb6f-2ce9-70f0-b717-5577bda68903", "title": "Insurance Voted First Why FSD 13 / 14 / 15 May Reprice the Entire Mobility Industry", "url": "https://liweinlp.com/13463", "published_at": "2026-01-25T21:41:08+00:00" }, { "id": "01a0bb6f-2ce9-70f0-b717-5577be0775c2", "title": "从 FSD 13 到 Unsupervised（F15）：自动驾驶如何穿透保险、监管与商业模式", "url": "https://liweinlp.com/13461", "published_at": "2026-01-25T21:11:10+00:00" }, { "id": "01a0bb6f-2ce9-70f0-b717-5577be79cd0b", "title": "自动驾驶监管过度，实质是一种间接杀人", "url": "https://liweinlp.com/13458", "published_at": "2026-01-25T18:31:21+00:00" }, { "id": "01a0bb6f-2ce9-70f0-b717-5577bf1c111d", "title": "多少牛人陷入大模型的认知茧房？", "url": "https://liweinlp.com/13455", "published_at": "2026-01-23T03:02:23+00:00" }, { "id": "01a0bb6f-2ce9-70f0-b717-5577bf5267d0", "title": "从 “Fake It” 到 “Vibe It”", "url": "https://liweinlp.com/13452", "published_at": "2026-01-18T21:28:47+00:00" }, { "id": "01a0bb6f-2ce9-70f0-b717-5577bfbc1532", "title": "耳机是你的贴身陪伴吗", "url": "https://liweinlp.com/13451", "published_at": "2026-01-13T09:19:55+00:00" }, { "id": "01a0bb6f-2ce9-70f0-b717-5577c0b15c0a", "title": "从open ai 主打的 耳后AI耳机谈起", "url": "https://liweinlp.com/13449", "published_at": "2026-01-13T09:04:46+00:00" }, { "id": "01a0bb6f-2ce9-70f0-b717-5577c114a2b4", "title": "全双工到天花板的豆包", "url": "https://liweinlp.com/13447", "published_at": "2026-01-03T14:34:34+00:00" }, { "id": "01a0bb6f-2ce9-70f0-b717-5577c204cc57", "title": "FSD + Grok：超人老司机，带着“实习导游”上路", "url": "https://liweinlp.com/13446", "published_at": "2026-01-03T03:00:39+00:00" }, { "id": "01a0bb6f-2ce9-70f0-b717-5577c2650440", "title": "梁文峰团队的 mHC 研究在做什么", "url": "https://liweinlp.com/13440", "published_at": "2026-01-02T09:11:32+00:00" }, { "id": "01a0bb6f-2ce9-70f0-b717-5577c2fc3bff", "title": "从“眼球 + SaaS”到“大模型商业”", "url": "https://liweinlp.com/13438", "published_at": "2025-12-18T17:00:43+00:00" }, { "id": "01a0bb6f-2ce9-70f0-b717-5577c34eea11", "title": "2025 AGI 编年史：从范式革命到商业黎明", "url": "https://liweinlp.com/13432", "published_at": "2025-12-18T14:07:00+00:00" }, { "id": "01a0bb6f-2ce9-70f0-b717-5577c418bf47", "title": "AI Reflections on 2025", "url": "https://liweinlp.com/13433", "published_at": "2025-12-18T13:07:07+00:00" }, { "id": "01a0bb6f-2ce9-70f0-b717-5577c4e03740", "title": "2025 年 AI 感怀", "url": "https://liweinlp.com/13430", "published_at": "2025-12-18T02:51:45+00:00" }, { "id": "01a0bb6f-2ce9-70f0-b717-5577c5cfbfeb", "title": "正常的模型反哺不会导致模型坍塌", "url": "https://liweinlp.com/13425", "published_at": "2025-12-16T03:12:33+00:00" }, { "id": "01a0bb72-de34-70d8-bfab-91184294435b", "title": "多模态进化论：从“看图说话”到“原生直觉”", "url": "https://liweinlp.com/13422", "published_at": "2025-12-15T19:13:36+00:00" }, { "id": "01a0bb72-de34-70d8-bfab-911843699a99", "title": "大模型训练的数据“炼金术”", "url": "https://liweinlp.com/13420", "published_at": "2025-12-15T07:28:36+00:00" }, { "id": "01a0bb72-de34-70d8-bfab-9118443f3ee6", "title": "再论大模型压缩的“有损”与“无损”", "url": "https://liweinlp.com/13414", "published_at": "2025-11-24T04:02:48+00:00" }, { "id": "01a0bb72-de34-70d8-bfab-9118445a261d", "title": "大模型是无损压缩还是有损压缩，李飞飞与伊利亚谁是对的？", "url": "https://liweinlp.com/13413", "published_at": "2025-11-24T02:56:02+00:00" }, { "id": "01a0bb72-de34-70d8-bfab-9118447a7357", "title": "视频语义检索怎么回事", "url": "https://liweinlp.com/13410", "published_at": "2025-11-18T17:49:38+00:00" }, { "id": "01a0bb72-de34-70d8-bfab-911845348593", "title": "GPT非监督学习到底怎么就学会了各种监督任务呢？", "url": "https://liweinlp.com/13407", "published_at": "2025-11-10T04:18:32+00:00" }, { "id": "01a0bb72-de34-70d8-bfab-9118461d97ce", "title": "The Chain Rule: The Mathematical Guarantee Behind Backpropagation", "url": "https://liweinlp.com/13405", "published_at": "2025-11-07T02:58:53+00:00" }, { "id": "01a0bb72-de34-70d8-bfab-911846e87bdf", "title": "Backpropagation: The Key to Deep Neural Networks", "url": "https://liweinlp.com/13404", "published_at": "2025-11-07T02:43:00+00:00" }, { "id": "01a0bb72-de34-70d8-bfab-911847a618d5", "title": "链式法则：反向传播能work的数学保证", "url": "https://liweinlp.com/13403", "published_at": "2025-11-07T00:55:42+00:00" }, { "id": "01a0bb72-de34-70d8-bfab-911847a852cf", "title": "反向传播：深层神经网络的钥匙", "url": "https://liweinlp.com/13402", "published_at": "2025-11-07T00:44:43+00:00" }, { "id": "01a0bb72-de34-70d8-bfab-9118488a9b19", "title": "CNN与RNN——让机器学会看与听", "url": "https://liweinlp.com/13401", "published_at": "2025-11-06T22:38:16+00:00" }, { "id": "01a0bb72-de34-70d8-bfab-911848a086ed", "title": "自学习是思想革命，Transformer是工程火箭", "url": "https://liweinlp.com/13400", "published_at": "2025-11-06T22:17:56+00:00" }, { "id": "01a0bb72-de34-70d8-bfab-9118490f0c72", "title": "从高级语言的基本逻辑装置到图灵机的编译", "url": "https://liweinlp.com/13392", "published_at": "2025-09-19T00:52:05+00:00" }, { "id": "01a0bb72-de34-70d8-bfab-9118495809c7", "title": "尼克讲座第二弹：语言=思维=智能=计算=图灵机？", "url": "https://liweinlp.com/13387", "published_at": "2025-09-19T00:25:49+00:00" }, { "id": "01a0bb72-de34-70d8-bfab-911849b39fb9", "title": "小科普：图灵机是怎么工作的？", "url": "https://liweinlp.com/13388", "published_at": "2025-09-18T23:57:00+00:00" }, { "id": "01a0bb72-de34-70d8-bfab-911849dc35cb", "title": "Multimodal tokens and the Case for Unified Modeling", "url": "https://liweinlp.com/13384", "published_at": "2025-09-15T03:54:20+00:00" }, { "id": "01a0bb72-de34-70d8-bfab-91184a62fcd3", "title": "谈谈大一统模型的万象token与统一建模的背后", "url": "https://liweinlp.com/13382", "published_at": "2025-09-15T03:20:57+00:00" }, { "id": "01a0bb72-de34-70d8-bfab-91184af6be89", "title": "老马的共产主义宣言", "url": "https://liweinlp.com/13381", "published_at": "2025-09-14T03:02:45+00:00" }, { "id": "01a0bb72-de34-70d8-bfab-91184bbe1d78", "title": "Neural Codec: Key Audio Techniques in the LLM Era", "url": "https://liweinlp.com/13379", "published_at": "2025-09-12T18:04:51+00:00" }, { "id": "01a0bb72-de34-70d8-bfab-91184c8f1eca", "title": "Breakthroughs in Speech Technology in the Era of Large Models: Ultra-Realism and Full Duplex", "url": "https://liweinlp.com/13377", "published_at": "2025-09-12T17:55:37+00:00" }, { "id": "01a0bb76-2a29-723b-a291-2e73d011b695", "title": "大模型时代的语音技术突破：超写实和全双工", "url": "https://liweinlp.com/13374", "published_at": "2025-09-12T17:19:01+00:00" }, { "id": "01a0bb76-2a29-723b-a291-2e73d0d6cfdc", "title": "说说神经 codec，大模型时代的音频技术要点", "url": "https://liweinlp.com/13372", "published_at": "2025-09-12T09:04:14+00:00" }, { "id": "01a0bb76-2a29-723b-a291-2e73d128fe7a", "title": "跨模态连接器范式：谷歌模型Flamingo回顾", "url": "https://liweinlp.com/13369", "published_at": "2025-09-03T01:16:34+00:00" }, { "id": "01a0bb76-2a29-723b-a291-2e73d12a4361", "title": "图文对齐的关键一跃：CLIP 回顾", "url": "https://liweinlp.com/13365", "published_at": "2025-09-02T03:18:53+00:00" }, { "id": "01a0bb76-2a29-723b-a291-2e73d1887939", "title": "Attention Collapse: The Misunderstood Truth About “Rank”", "url": "https://liweinlp.com/13364", "published_at": "2025-08-30T05:00:35+00:00" }, { "id": "01a0bb76-2a29-723b-a291-2e73d19656df", "title": "注意力塌缩：关于“秩”的误会与真相", "url": "https://liweinlp.com/13356", "published_at": "2025-08-30T01:06:42+00:00" }, { "id": "01a0bb76-2a29-723b-a291-2e73d1c01cc3", "title": "超长上下文，是能力突破，还是数字游戏？", "url": "https://liweinlp.com/13353", "published_at": "2025-08-29T06:32:45+00:00" }, { "id": "01a0bb76-2a29-723b-a291-2e73d2998979", "title": "BERT 双向 vs. GPT 单向与“低秩之虑”", "url": "https://liweinlp.com/13349", "published_at": "2025-08-28T01:57:23+00:00" }, { "id": "01a0bb76-2a29-723b-a291-2e73d30b4b41", "title": "自监督学习的两大模型，为什么GPT跑赢了BERT成为王者？", "url": "https://liweinlp.com/13346", "published_at": "2025-08-20T10:46:51+00:00" }, { "id": "01a0bb76-2a29-723b-a291-2e73d34e2a08", "title": "Is the World Material or Informational?", "url": "https://liweinlp.com/13343", "published_at": "2025-08-18T11:57:44+00:00" }, { "id": "01a0bb76-2a29-723b-a291-2e73d3f4c61a", "title": "“宇宙操作系统”：世界是怎么搭建的？", "url": "https://liweinlp.com/13341", "published_at": "2025-08-18T10:36:50+00:00" }, { "id": "01a0bb76-2a29-723b-a291-2e73d48630fc", "title": "AI 的威胁：不是恶意，而是作用链", "url": "https://liweinlp.com/13340", "published_at": "2025-08-18T05:45:00+00:00" }, { "id": "01a0bb76-2a29-723b-a291-2e73d4939d28", "title": "一个日常生活真实需求的 Deep Research 案例", "url": "https://liweinlp.com/13333", "published_at": "2025-08-08T20:13:29+00:00" }, { "id": "01a0bb76-2a29-723b-a291-2e73d4be8a41", "title": "老友访谈：AI对工作市场的影响", "url": "https://liweinlp.com/13329", "published_at": "2025-08-03T18:57:03+00:00" }, { "id": "01a0bb76-2a29-723b-a291-2e73d5a92a21", "title": "从 Suno 看 AIGC 艺术民主化大潮", "url": "https://liweinlp.com/13325", "published_at": "2025-08-02T17:42:13+00:00" }, { "id": "01a0bb76-2a29-723b-a291-2e73d5cbfb2f", "title": "狼来了，狼来了，“奇点”狼这次是真要来了吗？", "url": "https://liweinlp.com/13323", "published_at": "2025-07-31T18:51:18+00:00" }, { "id": "01a0bb76-2a29-723b-a291-2e73d6383771", "title": "notebookLM赋能：隐藏推理，大模型推理模型的新动向", "url": "https://liweinlp.com/13321", "published_at": "2025-07-31T02:28:35+00:00" }, { "id": "01a0bb76-2a29-723b-a291-2e73d7156e55", "title": "JJ：家庭四季", "url": "https://liweinlp.com/13317", "published_at": "2025-07-26T21:07:53+00:00" }, { "id": "01a0bb76-2a29-723b-a291-2e73d715adfe", "title": "Is Thinking Equal to Language?", "url": "https://liweinlp.com/13314", "published_at": "2025-07-25T01:28:51+00:00" }, { "id": "01a0bb76-2a29-723b-a291-2e73d80b079f", "title": "思维等于语言吗??", "url": "https://liweinlp.com/13308", "published_at": "2025-07-25T00:31:59+00:00" }, { "id": "01a0bb79-705e-7176-b479-e04f25d99447", "title": "Demystifying the misconception of \"Lossless Compression as Intelligence\"", "url": "https://liweinlp.com/13306", "published_at": "2025-07-06T18:16:45+00:00" }, { "id": "01a0bb79-705e-7176-b479-e04f26b2ab8b", "title": "破除“无损压缩即智能”的迷思", "url": "https://liweinlp.com/13274", "published_at": "2025-07-06T17:34:08+00:00" }, { "id": "01a0bb79-705e-7176-b479-e04f2759a4a5", "title": "GPT and the Art of Compression", "url": "https://liweinlp.com/13295", "published_at": "2025-07-05T18:20:21+00:00" }, { "id": "01a0bb79-705e-7176-b479-e04f2857dc9b", "title": "Efficiency vs. Reliability: The Compression Tightrope", "url": "https://liweinlp.com/13294", "published_at": "2025-07-05T06:01:54+00:00" }, { "id": "01a0bb79-705e-7176-b479-e04f293d397d", "title": "Arithmetic Coding for GPT’s Compression Engine", "url": "https://liweinlp.com/13293", "published_at": "2025-07-05T05:42:52+00:00" }, { "id": "01a0bb79-705e-7176-b479-e04f2947eda2", "title": "Navigating the Probability Universe with GPT", "url": "https://liweinlp.com/13292", "published_at": "2025-07-05T05:37:16+00:00" }, { "id": "01a0bb79-705e-7176-b479-e04f2a474323", "title": "Is GPT Compression Lossless or Lossy? The Truth Revealed", "url": "https://liweinlp.com/13291", "published_at": "2025-07-05T05:30:45+00:00" }, { "id": "01a0bb79-705e-7176-b479-e04f2aafbe46", "title": "GPT as a Cosmic Librarian: Unlocking Lossless Compression", "url": "https://liweinlp.com/13290", "published_at": "2025-07-05T05:18:32+00:00" }, { "id": "01a0bb79-705e-7176-b479-e04f2b6ab814", "title": "与尼克等老友唠大模型压缩理论", "url": "https://liweinlp.com/13288", "published_at": "2025-07-05T02:49:48+00:00" }, { "id": "01a0bb79-705e-7176-b479-e04f2c371cdd", "title": "GPT作为序列数据的无损压缩器", "url": "https://liweinlp.com/13277", "published_at": "2025-07-05T01:17:55+00:00" }, { "id": "01a0bb79-705e-7176-b479-e04f2d0cb266", "title": "要区分GPT训练中的压缩，与拿GPT当压缩器工具", "url": "https://liweinlp.com/13281", "published_at": "2025-07-04T21:15:13+00:00" }, { "id": "01a0bb79-705e-7176-b479-e04f2ddd2f47", "title": "信息论科普：GPT对给定序列无损压缩的最终区间", "url": "https://liweinlp.com/13279", "published_at": "2025-07-04T19:57:56+00:00" }, { "id": "01a0bb79-705e-7176-b479-e04f2ed155ae", "title": "GPT无损压缩小问答（3）：算术编码", "url": "https://liweinlp.com/13275", "published_at": "2025-07-04T19:27:38+00:00" }, { "id": "01a0bb79-705e-7176-b479-e04f2f55ed07", "title": "信息论科普：香农极限（Shannon Limit）", "url": "https://liweinlp.com/13276", "published_at": "2025-07-04T18:16:32+00:00" }, { "id": "01a0bb79-705e-7176-b479-e04f2ff042c1", "title": "GPT无损压缩小问答（2）：为什么说GPT是无损压缩？", "url": "https://liweinlp.com/13273", "published_at": "2025-07-04T16:43:33+00:00" }, { "id": "01a0bb79-705e-7176-b479-e04f300ffb89", "title": "GPT无损压缩小问答（1）： 高压缩率导致系统脆弱", "url": "https://liweinlp.com/13272", "published_at": "2025-07-04T16:20:48+00:00" }, { "id": "01a0bb79-705e-7176-b479-e04f30f934a4", "title": "像素值是“连续变量”，还是工程上的伪装？", "url": "https://liweinlp.com/13266", "published_at": "2025-06-21T13:52:38+00:00" }, { "id": "01a0bb79-705e-7176-b479-e04f31df2361", "title": "Yann LeCun 所鼓吹的「世界模型」与GPT+Diffusion有什么不同", "url": "https://liweinlp.com/13263", "published_at": "2025-06-12T14:01:28+00:00" }, { "id": "01a0bb79-705e-7176-b479-e04f32bb8f59", "title": "《立委关于大模型与AI的博客汇总》", "url": "https://liweinlp.com/13260", "published_at": "2025-06-08T01:40:33+00:00" }, { "id": "01a0bb79-705e-7176-b479-e04f338caf7a", "title": "详述AI科学家張祥雨的大模型访谈", "url": "https://liweinlp.com/13256", "published_at": "2025-06-05T05:16:03+00:00" } ] posts Claim your blog
Back to liweinlp.com
Blog · corpus.blog/blogs/liweinlp.com/posts

liweinlp.com

liweinlp.com

2026

新史记：CC外泄记original ↗
4 Apr 2026
关于Claude Code 泄漏事件的愚人节思考original ↗
2 Apr 2026
What the Claude Code TypeScript Leak Really Revealedoriginal ↗
2 Apr 2026
When Agents Become the Default Gateway, Will the Operating System Be Rewritten?original ↗
13 Feb 2026
When Agents Become the Default Gateway, Does the App Store Model Collapse?original ↗
13 Feb 2026
The Great Software Shake-Up of the Agent Era — Starting with OpenClaworiginal ↗
13 Feb 2026
Some Basic Agentic AI terminologyoriginal ↗
13 Feb 2026
OpenClaw as a case study of the coming Agentic AI eraoriginal ↗
12 Feb 2026
Agent 时代的临界点：谈谈 OpenClaw 的安全隐患original ↗
12 Feb 2026
Agent 时代的一些术语澄清original ↗
12 Feb 2026
当 Agent 成为默认入口，App Store 模式是否崩塌？original ↗
12 Feb 2026
当 Agent 成为默认入口，操作系统会不会被重写？original ↗
12 Feb 2026
Agent 时代的软件产业大洗牌：从 OpenClaw谈起original ↗
12 Feb 2026
2026年智能体范式大爆发：从认知幻象到工业化协同original ↗
11 Feb 2026
腾讯科技春节访谈，Agent 这一年：沸沸扬扬之后original ↗
10 Feb 2026
多模态大模型输入信号的离散化original ↗
1 Feb 2026
置顶:《立委关于大模型与AI的科学网博客汇总》original ↗
29 Jan 2026
The Removal of Autopilot: A Misjudgment of Trust, Pricing Power, and Timingoriginal ↗
28 Jan 2026
Autopilot 被剥离: 一次关于信任与定价权的误判original ↗
28 Jan 2026
自动驾驶已经解决了，但我们还没准备好告别驾驶original ↗
28 Jan 2026
How FSD Quietly Took Control of Pricing Poweroriginal ↗
26 Jan 2026
If Robotaxi Fails, This Is Where It Will Failoriginal ↗
26 Jan 2026
FSD 会拯救“最不被保险欢迎的人”original ↗
26 Jan 2026
保险降价，是自动驾驶第一次“自证盈利模型”original ↗
26 Jan 2026
如果 FSD 真的会失败，特斯拉最可能栽在哪里？original ↗
25 Jan 2026
Insurance Voted First Why FSD 13 / 14 / 15 May Reprice the Entire Mobility Industryoriginal ↗
25 Jan 2026
从 FSD 13 到 Unsupervised（F15）：自动驾驶如何穿透保险、监管与商业模式original ↗
25 Jan 2026
自动驾驶监管过度，实质是一种间接杀人original ↗
25 Jan 2026
多少牛人陷入大模型的认知茧房？original ↗
23 Jan 2026
从 “Fake It” 到 “Vibe It”original ↗
18 Jan 2026
耳机是你的贴身陪伴吗original ↗
13 Jan 2026
从open ai 主打的 耳后AI耳机谈起original ↗
13 Jan 2026
全双工到天花板的豆包original ↗
3 Jan 2026
FSD + Grok：超人老司机，带着“实习导游”上路original ↗
3 Jan 2026
梁文峰团队的 mHC 研究在做什么original ↗
2 Jan 2026

2025

从“眼球 + SaaS”到“大模型商业”original ↗
18 Dec 2025
2025 AGI 编年史：从范式革命到商业黎明original ↗
18 Dec 2025
AI Reflections on 2025original ↗
18 Dec 2025
2025 年 AI 感怀original ↗
18 Dec 2025
正常的模型反哺不会导致模型坍塌original ↗
16 Dec 2025
多模态进化论：从“看图说话”到“原生直觉”original ↗
15 Dec 2025
大模型训练的数据“炼金术”original ↗
15 Dec 2025
再论大模型压缩的“有损”与“无损”original ↗
24 Nov 2025
大模型是无损压缩还是有损压缩，李飞飞与伊利亚谁是对的？original ↗
24 Nov 2025
视频语义检索怎么回事original ↗
18 Nov 2025
GPT非监督学习到底怎么就学会了各种监督任务呢？original ↗
10 Nov 2025
The Chain Rule: The Mathematical Guarantee Behind Backpropagationoriginal ↗
7 Nov 2025
Backpropagation: The Key to Deep Neural Networksoriginal ↗
7 Nov 2025
链式法则：反向传播能work的数学保证original ↗
7 Nov 2025
反向传播：深层神经网络的钥匙original ↗
7 Nov 2025
CNN与RNN——让机器学会看与听original ↗
6 Nov 2025
自学习是思想革命，Transformer是工程火箭original ↗
6 Nov 2025
从高级语言的基本逻辑装置到图灵机的编译original ↗
19 Sept 2025
尼克讲座第二弹：语言=思维=智能=计算=图灵机？original ↗
19 Sept 2025
小科普：图灵机是怎么工作的？original ↗
18 Sept 2025
Multimodal tokens and the Case for Unified Modelingoriginal ↗
15 Sept 2025
谈谈大一统模型的万象token与统一建模的背后original ↗
15 Sept 2025
老马的共产主义宣言original ↗
14 Sept 2025
Neural Codec: Key Audio Techniques in the LLM Eraoriginal ↗
12 Sept 2025
Breakthroughs in Speech Technology in the Era of Large Models: Ultra-Realism and Full Duplexoriginal ↗
12 Sept 2025
大模型时代的语音技术突破：超写实和全双工original ↗
12 Sept 2025
说说神经 codec，大模型时代的音频技术要点original ↗
12 Sept 2025
跨模态连接器范式：谷歌模型Flamingo回顾original ↗
3 Sept 2025
图文对齐的关键一跃：CLIP 回顾original ↗
2 Sept 2025
Attention Collapse: The Misunderstood Truth About “Rank”original ↗
30 Aug 2025
注意力塌缩：关于“秩”的误会与真相original ↗
30 Aug 2025
超长上下文，是能力突破，还是数字游戏？original ↗
29 Aug 2025
BERT 双向 vs. GPT 单向与“低秩之虑”original ↗
28 Aug 2025
自监督学习的两大模型，为什么GPT跑赢了BERT成为王者？original ↗
20 Aug 2025
Is the World Material or Informational?original ↗
18 Aug 2025
“宇宙操作系统”：世界是怎么搭建的？original ↗
18 Aug 2025
AI 的威胁：不是恶意，而是作用链original ↗
18 Aug 2025
一个日常生活真实需求的 Deep Research 案例original ↗
8 Aug 2025
老友访谈：AI对工作市场的影响original ↗
3 Aug 2025
从 Suno 看 AIGC 艺术民主化大潮original ↗
2 Aug 2025
狼来了，狼来了，“奇点”狼这次是真要来了吗？original ↗
31 Jul 2025
notebookLM赋能：隐藏推理，大模型推理模型的新动向original ↗
31 Jul 2025
JJ：家庭四季original ↗
26 Jul 2025
Is Thinking Equal to Language?original ↗
25 Jul 2025
思维等于语言吗??original ↗
25 Jul 2025
Demystifying the misconception of "Lossless Compression as Intelligence"original ↗
6 Jul 2025
破除“无损压缩即智能”的迷思original ↗
6 Jul 2025
GPT and the Art of Compressionoriginal ↗
5 Jul 2025
Efficiency vs. Reliability: The Compression Tightropeoriginal ↗
5 Jul 2025
Arithmetic Coding for GPT’s Compression Engineoriginal ↗
5 Jul 2025
Navigating the Probability Universe with GPToriginal ↗
5 Jul 2025
Is GPT Compression Lossless or Lossy? The Truth Revealedoriginal ↗
5 Jul 2025
GPT as a Cosmic Librarian: Unlocking Lossless Compressionoriginal ↗
5 Jul 2025
与尼克等老友唠大模型压缩理论original ↗
5 Jul 2025
GPT作为序列数据的无损压缩器original ↗
5 Jul 2025
要区分GPT训练中的压缩，与拿GPT当压缩器工具original ↗
4 Jul 2025
信息论科普：GPT对给定序列无损压缩的最终区间original ↗
4 Jul 2025
GPT无损压缩小问答（3）：算术编码original ↗
4 Jul 2025
信息论科普：香农极限（Shannon Limit）original ↗
4 Jul 2025
GPT无损压缩小问答（2）：为什么说GPT是无损压缩？original ↗
4 Jul 2025
GPT无损压缩小问答（1）： 高压缩率导致系统脆弱original ↗
4 Jul 2025
像素值是“连续变量”，还是工程上的伪装？original ↗
21 Jun 2025
Yann LeCun 所鼓吹的「世界模型」与GPT+Diffusion有什么不同original ↗
12 Jun 2025
《立委关于大模型与AI的博客汇总》original ↗
8 Jun 2025
详述AI科学家張祥雨的大模型访谈original ↗
5 Jun 2025
Previous Next