corpus.blog
41554 blogs · [ { "id": "01a0bb7d-91ea-7146-a683-d4fd192fc181", "title": "LLM的后训练强化学习是怎么工作的", "url": "https://liweinlp.com/13252", "published_at": "2025-06-01T19:23:03+00:00" }, { "id": "01a0bb7d-91ea-7146-a683-d4fd19640c65", "title": "大模型科普：探秘莎翁风格的诞生之旅（无代码版）", "url": "https://liweinlp.com/13248", "published_at": "2025-05-31T13:19:08+00:00" }, { "id": "01a0bb7d-91ea-7146-a683-d4fd19ee75d7", "title": "揭秘GPT内核之四", "url": "https://liweinlp.com/13243", "published_at": "2025-05-31T10:33:04+00:00" }, { "id": "01a0bb7d-91ea-7146-a683-d4fd1a6657f7", "title": "How GPT Works: A Shakespearean Text Generator", "url": "https://liweinlp.com/13241", "published_at": "2025-05-30T19:55:19+00:00" }, { "id": "01a0bb7d-91ea-7146-a683-d4fd1a7ae2df", "title": "从零实现莎士比亚风 GPT科普解说", "url": "https://liweinlp.com/13239", "published_at": "2025-05-30T19:50:54+00:00" }, { "id": "01a0bb7d-91ea-7146-a683-d4fd1b1e1aec", "title": "从0实现并理解GPT", "url": "https://liweinlp.com/13234", "published_at": "2025-05-28T15:21:13+00:00" }, { "id": "01a0bb7d-91ea-7146-a683-d4fd1babf7ab", "title": "Decoding the New EMPO Reasoning Paradigm", "url": "https://liweinlp.com/13233", "published_at": "2025-05-27T06:01:36+00:00" }, { "id": "01a0bb7d-91ea-7146-a683-d4fd1c8a744c", "title": "解读EMPO全程无监督推理新范式", "url": "https://liweinlp.com/13230", "published_at": "2025-05-27T05:03:04+00:00" }, { "id": "01a0bb7d-91eb-72e9-83f8-ee57c34495d3", "title": "MeanFlow: AI图像生成的降维打击", "url": "https://liweinlp.com/13224", "published_at": "2025-05-22T11:18:47+00:00" }, { "id": "01a0bb7d-91eb-72e9-83f8-ee57c3eb6f9d", "title": "Q&A on NLP: Chapter I Natural Language and Linguistic Form", "url": "https://liweinlp.com/13211", "published_at": "2025-05-09T21:34:30+00:00" }, { "id": "01a0bb7d-91eb-72e9-83f8-ee57c49a490f", "title": "Prelude: Origins", "url": "https://liweinlp.com/13210", "published_at": "2025-05-09T05:50:59+00:00" }, { "id": "01a0bb7d-91eb-72e9-83f8-ee57c58b108e", "title": "Preface for \"Q&A on NLP\"", "url": "https://liweinlp.com/13206", "published_at": "2025-05-09T05:05:06+00:00" }, { "id": "01a0bb7d-91eb-72e9-83f8-ee57c6553644", "title": "关于模型蒸馏和 KL散度的问答", "url": "https://liweinlp.com/13201", "published_at": "2025-05-07T05:49:14+00:00" }, { "id": "01a0bb7d-91eb-72e9-83f8-ee57c65e900b", "title": "A Comparative Review of Autoregressive and Diffusion Models for Video Generation", "url": "https://liweinlp.com/13198", "published_at": "2025-05-02T07:01:04+00:00" }, { "id": "01a0bb7d-91eb-72e9-83f8-ee57c6a95670", "title": "Unveiling the Two \"Superpowers\" Behind AI Video Creation", "url": "https://liweinlp.com/13195", "published_at": "2025-05-02T04:11:42+00:00" }, { "id": "01a0bb7d-91eb-72e9-83f8-ee57c6d3a319", "title": "立委科普：揭秘AI创作视频的两种“神功”", "url": "https://liweinlp.com/13192", "published_at": "2025-05-01T22:19:36+00:00" }, { "id": "01a0bb7d-91eb-72e9-83f8-ee57c79f3a02", "title": "非量化自回归视频生成模型NOVA的技术路线", "url": "https://liweinlp.com/13183", "published_at": "2025-05-01T05:38:26+00:00" }, { "id": "01a0bb7d-91eb-72e9-83f8-ee57c870f814", "title": "生成式AI的两条视频生成路线", "url": "https://liweinlp.com/13180", "published_at": "2025-04-30T03:15:46+00:00" }, { "id": "01a0bb7d-91eb-72e9-83f8-ee57c8c7a2f5", "title": "MCP: From Flashy Boom to Real Usability — A Technical Deep Dive", "url": "https://liweinlp.com/13176", "published_at": "2025-04-26T08:45:08+00:00" }, { "id": "01a0bb7d-91eb-72e9-83f8-ee57c991b7aa", "title": "MCP：从“伪繁荣”到可落地的进化路线", "url": "https://liweinlp.com/13175", "published_at": "2025-04-26T08:30:52+00:00" }, { "id": "01a0bb80-9cec-70c9-8532-520294ab58cf", "title": "Silicon Valley Night: A Foxy Encounter", "url": "https://liweinlp.com/13173", "published_at": "2025-04-23T02:57:15+00:00" }, { "id": "01a0bb80-9cec-70c9-8532-520294f1c221", "title": "《硅谷夜记：艳遇》", "url": "https://liweinlp.com/13165", "published_at": "2025-04-22T21:13:15+00:00" }, { "id": "01a0bb80-9cec-70c9-8532-5202958f4a80", "title": "短剧：黄石的低语 (Whispers of Yellowstone)", "url": "https://liweinlp.com/13158", "published_at": "2025-04-03T17:39:16+00:00" }, { "id": "01a0bb80-9cec-70c9-8532-520295ea47e8", "title": "breaking news：黄石起义：熊的宣言", "url": "https://liweinlp.com/13155", "published_at": "2025-04-03T17:25:23+00:00" }, { "id": "01a0bb80-9cec-70c9-8532-520296bf958f", "title": "自传体小说《刀锋人生：百年缝合》（3/3）", "url": "https://liweinlp.com/13154", "published_at": "2025-04-03T10:05:47+00:00" }, { "id": "01a0bb80-9cec-70c9-8532-52029753748c", "title": "自传体小说《刀锋人生：百年缝合》（2）", "url": "https://liweinlp.com/13152", "published_at": "2025-04-03T05:11:10+00:00" }, { "id": "01a0bb80-9cec-70c9-8532-5202983c4e30", "title": "自传体小说：《刀锋人生：百年缝合》", "url": "https://liweinlp.com/13149", "published_at": "2025-04-02T23:17:23+00:00" }, { "id": "01a0bb80-9cec-70c9-8532-520298572214", "title": "The Scalpel’s Edge: A Life Stitched Through a Century (3）", "url": "https://liweinlp.com/13147", "published_at": "2025-04-02T22:57:19+00:00" }, { "id": "01a0bb80-9cec-70c9-8532-5202985f8416", "title": "The Scalpel’s Edge: A Life Stitched Through a Century (2）", "url": "https://liweinlp.com/13148", "published_at": "2025-04-02T22:29:46+00:00" }, { "id": "01a0bb80-9cec-70c9-8532-52029875d463", "title": "The Scalpel’s Edge: A Life Stitched Through a Century (自传体小说）", "url": "https://liweinlp.com/13145", "published_at": "2025-04-02T21:49:11+00:00" }, { "id": "01a0bb80-9cec-70c9-8532-52029880b2b3", "title": "CHAPTER 15: RECENT GATHERING SPEECHES", "url": "https://liweinlp.com/13143", "published_at": "2025-04-02T17:44:03+00:00" }, { "id": "01a0bb80-9cec-70c9-8532-52029898337d", "title": "CHAPTER 16: THE LI FAMILY VALUES", "url": "https://liweinlp.com/13142", "published_at": "2025-04-02T17:43:03+00:00" }, { "id": "01a0bb80-9cec-70c9-8532-520298bf692f", "title": "CHAPTER 7: SEASONS OF WIND AND RAIN (1)", "url": "https://liweinlp.com/13138", "published_at": "2025-04-02T17:38:40+00:00" }, { "id": "01a0bb80-9cec-70c9-8532-520299628203", "title": "CHAPTER 5: SEASONS OF WIND AND RAIN", "url": "https://liweinlp.com/13137", "published_at": "2025-04-02T17:37:32+00:00" }, { "id": "01a0bb80-9cec-70c9-8532-520299b3b28f", "title": "CHAPTER 4: THE BURNING SUNSET GLOW", "url": "https://liweinlp.com/13136", "published_at": "2025-04-02T17:36:43+00:00" }, { "id": "01a0bb80-9cec-70c9-8532-520299b4776b", "title": "CHAPTER 3: MY SURGICAL CAREER—OUTSIDE THE HOSPITAL", "url": "https://liweinlp.com/13135", "published_at": "2025-04-02T17:36:03+00:00" }, { "id": "01a0bb80-9cec-70c9-8532-52029a475e22", "title": null, "url": "https://liweinlp.com/13134", "published_at": "2025-04-02T17:35:26+00:00" }, { "id": "01a0bb80-9cec-70c9-8532-52029a9bc10c", "title": "CHAPTER 11: CHRONICLES OF LIFE-AND-DEATH EXPERIENCES", "url": "https://liweinlp.com/13133", "published_at": "2025-04-02T17:34:32+00:00" }, { "id": "01a0bb80-9cec-70c9-8532-52029b8df0a7", "title": "CHAPTER 10: MY LIFE WITH THE 127TH HOSPITAL", "url": "https://liweinlp.com/13132", "published_at": "2025-04-02T17:32:56+00:00" }, { "id": "01a0bb80-9cec-70c9-8532-52029c633102", "title": "CHAPTER 13: ANOTHER COLLEGE ENTRANCE EXAMINATION APPROACHES", "url": "https://liweinlp.com/13129", "published_at": "2025-04-02T17:29:13+00:00" }, { "id": "01a0bb8e-38cc-71ef-b1cc-fce6ab6c4981", "title": "CHAPTER 14: SWEET – TANIA'S BRILLIANT LIFE", "url": "https://liweinlp.com/13127", "published_at": "2025-04-02T15:39:49+00:00" }, { "id": "01a0bb8e-38cc-71ef-b1cc-fce6ac5b719e", "title": "中文分词的前世今生", "url": "https://liweinlp.com/13123", "published_at": "2025-03-29T20:40:57+00:00" }, { "id": "01a0bb8e-38cc-71ef-b1cc-fce6ad293bf5", "title": "大模型如何解锁AI各种任务成为通用引擎的？", "url": "https://liweinlp.com/13113", "published_at": "2025-03-28T04:22:13+00:00" }, { "id": "01a0bb8e-38cc-71ef-b1cc-fce6ad65e4af", "title": "Grok：大模型为什么要超大数据？（4o配图）", "url": "https://liweinlp.com/13108", "published_at": "2025-03-27T05:28:03+00:00" }, { "id": "01a0bb8e-38cc-71ef-b1cc-fce6adb51b29", "title": "Grok: 大力出奇迹的背后", "url": "https://liweinlp.com/13107", "published_at": "2025-03-27T04:34:37+00:00" }, { "id": "01a0bb8e-38cc-71ef-b1cc-fce6ae1782b6", "title": "Grok：超大数据的大模型为何能收敛？", "url": "https://liweinlp.com/13099", "published_at": "2025-03-26T22:14:01+00:00" }, { "id": "01a0bb8e-38cc-71ef-b1cc-fce6aec2b13f", "title": "《“蜜蜂巢”里的子弹：JFK档案解密后》", "url": "https://liweinlp.com/13098", "published_at": "2025-03-24T19:57:09+00:00" }, { "id": "01a0bb8e-38cc-71ef-b1cc-fce6af1ccc5a", "title": "2025年肯尼迪遇刺档案解密揭示的惊人真相", "url": "https://liweinlp.com/13097", "published_at": "2025-03-24T19:40:35+00:00" }, { "id": "01a0bb8e-38cc-71ef-b1cc-fce6af3c8df4", "title": "2025年肯尼迪遇刺案档案解密", "url": "https://liweinlp.com/13095", "published_at": "2025-03-24T19:34:20+00:00" }, { "id": "01a0bb8e-38cc-71ef-b1cc-fce6afc90b97", "title": "Gemini Deep Research：用“Logits Lens”洞察神经网络的奥秘", "url": "https://liweinlp.com/13092", "published_at": "2025-03-22T20:33:30+00:00" }, { "id": "01a0bb8e-38cc-71ef-b1cc-fce6b0a8a7ff", "title": "03 deep research： Challenges and Prospects of Advanced Reasoning LLMs", "url": "https://liweinlp.com/13090", "published_at": "2025-03-20T15:47:41+00:00" }, { "id": "01a0bb8e-38cc-71ef-b1cc-fce6b155ad48", "title": "Sonnet3.7: 推理大模型的挑战与前景（图文版）", "url": "https://liweinlp.com/13082", "published_at": "2025-03-20T09:45:50+00:00" }, { "id": "01a0bb8e-38cc-71ef-b1cc-fce6b22ba0ef", "title": "Gemini deep research podcast： Challenges and Prospects of Advanced Reasoning LLMs", "url": "https://liweinlp.com/13080", "published_at": "2025-03-20T08:22:29+00:00" }, { "id": "01a0bb8e-38cc-71ef-b1cc-fce6b3246454", "title": "数学圆舞曲：欧拉恒等式（配乐诗朗诵）", "url": "https://liweinlp.com/13076", "published_at": "2025-03-19T19:32:13+00:00" }, { "id": "01a0bb8e-38cc-71ef-b1cc-fce6b3bb68c5", "title": "检索增强（RAG）与窗口数据的互补性 (图文版）", "url": "https://liweinlp.com/13074", "published_at": "2025-03-19T18:09:26+00:00" }, { "id": "01a0bb8e-38cc-71ef-b1cc-fce6b3d9d084", "title": "检索增强（RAG）与窗口数据的互补性", "url": "https://liweinlp.com/13072", "published_at": "2025-03-19T17:19:29+00:00" }, { "id": "01a0bb8e-38cc-71ef-b1cc-fce6b49629a9", "title": "人类 vs 恐龙：一场关于“不作不死”的滑稽短剧", "url": "https://liweinlp.com/13071", "published_at": "2025-03-16T16:52:51+00:00" }, { "id": "01a0bb8e-38cc-71ef-b1cc-fce6b5137807", "title": "deep research: 最新颈椎病手术指征与治疗概览", "url": "https://liweinlp.com/13067", "published_at": "2025-03-15T03:11:35+00:00" }, { "id": "01a0bb8e-38cc-71ef-b1cc-fce6b56f479f", "title": "关于颈椎病，大模型医疗建议靠谱吗？", "url": "https://liweinlp.com/13065", "published_at": "2025-03-14T22:14:04+00:00" }, { "id": "01a0bb8e-38cc-71ef-b1cc-fce6b6562515", "title": "给奶奶讲一下AI最新物种“大模型代理”", "url": "https://liweinlp.com/13064", "published_at": "2025-03-12T19:11:05+00:00" }, { "id": "01a0bb91-1c37-711f-8c3d-6913e6d7b964", "title": "Decoding LLM-native Agents: Bridging Compilation and Interpretation in AI", "url": "https://liweinlp.com/13063", "published_at": "2025-03-12T18:36:30+00:00" }, { "id": "01a0bb91-1c37-711f-8c3d-6913e6d7d710", "title": "Xiao Hong (Red): The Man Behind the Autonomus Genral Agent Manus", "url": "https://liweinlp.com/13061", "published_at": "2025-03-10T20:38:10+00:00" }, { "id": "01a0bb91-1c37-711f-8c3d-6913e7472613", "title": "The Agent Era: The Contemporary Evolution from Chatbots to Digital Agents", "url": "https://liweinlp.com/13060", "published_at": "2025-03-10T20:09:29+00:00" }, { "id": "01a0bb91-1c37-711f-8c3d-6913e7e1dc8d", "title": "o3 deep research: 智能体的分野和演进", "url": "https://liweinlp.com/13054", "published_at": "2025-03-10T09:53:42+00:00" }, { "id": "01a0bb91-1c37-711f-8c3d-6913e805148a", "title": "肖弘与Manus：AI Agent 的实战方法论", "url": "https://liweinlp.com/13057", "published_at": "2025-03-10T00:22:37+00:00" }, { "id": "01a0bb91-1c37-711f-8c3d-6913e834b031", "title": "Xiao Hong Red：肖弘其人", "url": "https://liweinlp.com/13055", "published_at": "2025-03-09T23:01:21+00:00" }, { "id": "01a0bb91-1c37-711f-8c3d-6913e89a49c9", "title": "万字长文解析 LLM-native Agent 及其混合计算方式", "url": "https://liweinlp.com/13051", "published_at": "2025-03-09T03:10:30+00:00" }, { "id": "01a0bb91-1c37-711f-8c3d-6913e8d1ffd8", "title": "o3 deep research: LLM 驱动的 Agent 综述", "url": "https://liweinlp.com/13046", "published_at": "2025-03-07T19:42:54+00:00" }, { "id": "01a0bb91-1c37-711f-8c3d-6913e93ce1b8", "title": "Agent：数字代理的崛起与未来", "url": "https://liweinlp.com/13045", "published_at": "2025-03-07T19:38:56+00:00" }, { "id": "01a0bb91-1c37-711f-8c3d-6913ea0446f5", "title": "Agent元年：从聊天机器人到数字员工的当代进化史", "url": "https://liweinlp.com/13043", "published_at": "2025-03-07T18:29:46+00:00" }, { "id": "01a0bb91-1c37-711f-8c3d-6913ea308357", "title": "【外一篇：推理范式演进中的概念】", "url": "https://liweinlp.com/13040", "published_at": "2025-03-06T17:52:17+00:00" }, { "id": "01a0bb91-1c37-711f-8c3d-6913eab89ad3", "title": "生成式AI学习中容易混淆的几个术语", "url": "https://liweinlp.com/13038", "published_at": "2025-03-05T18:52:23+00:00" }, { "id": "01a0bb91-1c37-711f-8c3d-6913eb603a91", "title": "思维链是大模型的符号神助攻", "url": "https://liweinlp.com/13034", "published_at": "2025-03-04T19:50:35+00:00" }, { "id": "01a0bb91-1c37-711f-8c3d-6913ebb72b42", "title": "再谈自然模态数据是高维空间的低维流形", "url": "https://liweinlp.com/13029", "published_at": "2025-03-04T00:03:21+00:00" }, { "id": "01a0bb91-1c37-711f-8c3d-6913ebbd4417", "title": "The Three-Stage Scaling Laws Large Language Models", "url": "https://liweinlp.com/13025", "published_at": "2025-03-03T07:00:42+00:00" }, { "id": "01a0bb91-1c37-711f-8c3d-6913ec98a2e9", "title": "大模型三阶段的 scaling laws 接力赛", "url": "https://liweinlp.com/13022", "published_at": "2025-03-03T02:58:04+00:00" }, { "id": "01a0bb91-1c37-711f-8c3d-6913eccc9bbb", "title": "Fundamental Limitations of Deep Learning: Origins in Data-Driven Methodology", "url": "https://liweinlp.com/13020", "published_at": "2025-03-02T18:40:48+00:00" }, { "id": "01a0bb91-1c37-711f-8c3d-6913ecd6290a", "title": "深度学习的局限性研究综述", "url": "https://liweinlp.com/13014", "published_at": "2025-03-02T05:39:40+00:00" }, { "id": "01a0bb91-1c37-711f-8c3d-6913ed3e0d70", "title": "o3 deep research: 深度学习局限性研究报告", "url": "https://liweinlp.com/13015", "published_at": "2025-03-02T04:46:56+00:00" }, { "id": "01a0bb91-1c37-711f-8c3d-6913ed635183", "title": "深度学习的基石：多层感知机", "url": "https://liweinlp.com/13012", "published_at": "2025-02-28T19:26:30+00:00" }, { "id": "01a0bb9d-e7b7-7050-96c7-340c3b642275", "title": "左脚踩右脚可以飞吗，谈交替使用监督微调和强化学习的后训练", "url": "https://liweinlp.com/13011", "published_at": "2025-02-27T21:16:18+00:00" }, { "id": "01a0bb9d-e7b7-7050-96c7-340c3b7c485d", "title": "o3 Deep Research: DeepSeek R1 多阶段训练流程问答解析", "url": "https://liweinlp.com/13008", "published_at": "2025-02-27T20:05:11+00:00" }, { "id": "01a0bb9d-e7b7-7050-96c7-340c3c120529", "title": "Dilemma of RPA and Early-Stage LLM Co-pilot Entrepreneurs in the Age of Agent Tsunami", "url": "https://liweinlp.com/13006", "published_at": "2025-02-27T05:06:23+00:00" }, { "id": "01a0bb9d-e7b7-7050-96c7-340c3c7c4e89", "title": "RPA 赛道与大模型Copilots早期创业者的困局", "url": "https://liweinlp.com/13003", "published_at": "2025-02-26T19:41:42+00:00" }, { "id": "01a0bb9d-e7b7-7050-96c7-340c3d755943", "title": "Understanding the Division of Labor Among Q, K, V in Self-Attention Mechanism", "url": "https://liweinlp.com/13000", "published_at": "2025-02-25T14:08:13+00:00" }, { "id": "01a0bb9d-e7b7-7050-96c7-340c3d90efe0", "title": "Linguists Should Find Self-Attention Intuitively Familiar", "url": "https://liweinlp.com/12999", "published_at": "2025-02-24T18:11:22+00:00" }, { "id": "01a0bb9d-e7b7-7050-96c7-340c3de78543", "title": "语言学家应该很容易理解自注意力机制", "url": "https://liweinlp.com/12994", "published_at": "2025-02-24T09:28:41+00:00" }, { "id": "01a0bb9d-e7b7-7050-96c7-340c3ecbcfe1", "title": "Introduction to Transformer and Its Attention Mechanism", "url": "https://liweinlp.com/12987", "published_at": "2025-02-23T22:39:33+00:00" }, { "id": "01a0bb9d-e7b7-7050-96c7-340c3f2f00fc", "title": "Has Symbolism Been Sidelined for Too Long? Could Neural LLM terminate AI?", "url": "https://liweinlp.com/12991", "published_at": "2025-02-23T17:57:16+00:00" }, { "id": "01a0bb9d-e7b7-7050-96c7-340c3f30f7ad", "title": "符号主义被打入冷宫太久了，难道神经是AI的终结者吗？", "url": "https://liweinlp.com/12989", "published_at": "2025-02-23T17:17:03+00:00" }, { "id": "01a0bb9d-e7b7-7050-96c7-340c40271199", "title": "Transformer 和注意力机制简介", "url": "https://liweinlp.com/12983", "published_at": "2025-02-20T20:49:46+00:00" }, { "id": "01a0bb9d-e7b7-7050-96c7-340c4042d7c3", "title": "立委科普：如何理解自注意力机制中的QKV分工？", "url": "https://liweinlp.com/12984", "published_at": "2025-02-20T19:39:35+00:00" }, { "id": "01a0bb9d-e7b7-7050-96c7-340c408688b6", "title": "True Story Behind DeepSeek's Success: AI Learning to Think Slowly Without Human Supervision", "url": "https://liweinlp.com/12977", "published_at": "2025-02-15T03:14:46+00:00" }, { "id": "01a0bb9d-e7b7-7050-96c7-340c40d1ec75", "title": "DeepSeek爆火真相：不靠“人盯”， 让AI自己学会慢思考", "url": "https://liweinlp.com/12975", "published_at": "2025-02-15T03:00:00+00:00" }, { "id": "01a0bb9d-e7b7-7050-96c7-340c418625e0", "title": "Does the New Reasoning Paradigm (Query+CoT+Answer) Support a New Scaling Law?", "url": "https://liweinlp.com/12973", "published_at": "2025-02-14T06:22:45+00:00" }, { "id": "01a0bb9d-e7b7-7050-96c7-340c4207149d", "title": "Technical Deep Dive: Understanding DeepSeek R1's Reasoning Mechanism in Production", "url": "https://liweinlp.com/12972", "published_at": "2025-02-14T06:12:55+00:00" }, { "id": "01a0bb9d-e7b7-7050-96c7-340c42dc2369", "title": "DeepSeek 笔记：推理新范式 query+cot+answer 支持新的 scaling law 吗？", "url": "https://liweinlp.com/12968", "published_at": "2025-02-13T20:00:52+00:00" }, { "id": "01a0bb9d-e7b7-7050-96c7-340c4354175f", "title": "DeepSeek 笔记：R1 部署阶段的推理机制", "url": "https://liweinlp.com/12967", "published_at": "2025-02-13T19:14:07+00:00" }, { "id": "01a0bb9d-e7b7-7050-96c7-340c443b5c17", "title": "Hallucinations in AI: Bug or Feature? A Deep Dive into DeepSeek-R1", "url": "https://liweinlp.com/12959", "published_at": "2025-02-09T06:29:43+00:00" }, { "id": "01a0bb9d-e7b7-7050-96c7-340c446678f8", "title": "从R1幻觉谈起，大模型幻觉是缺陷还是创意火花？", "url": "https://liweinlp.com/12961", "published_at": "2025-02-08T21:12:48+00:00" } ] posts Claim your blog
Back to liweinlp.com
Blog · corpus.blog/blogs/liweinlp.com/posts

liweinlp.com

liweinlp.com

2025

LLM的后训练强化学习是怎么工作的original ↗
1 Jun 2025
大模型科普：探秘莎翁风格的诞生之旅（无代码版）original ↗
31 May 2025
揭秘GPT内核之四original ↗
31 May 2025
How GPT Works: A Shakespearean Text Generatororiginal ↗
30 May 2025
从零实现莎士比亚风 GPT科普解说original ↗
30 May 2025
从0实现并理解GPToriginal ↗
28 May 2025
Decoding the New EMPO Reasoning Paradigmoriginal ↗
27 May 2025
解读EMPO全程无监督推理新范式original ↗
27 May 2025
MeanFlow: AI图像生成的降维打击original ↗
22 May 2025
Q&A on NLP: Chapter I Natural Language and Linguistic Formoriginal ↗
9 May 2025
Prelude: Originsoriginal ↗
9 May 2025
Preface for "Q&A on NLP"original ↗
9 May 2025
关于模型蒸馏和 KL散度的问答original ↗
7 May 2025
A Comparative Review of Autoregressive and Diffusion Models for Video Generationoriginal ↗
2 May 2025
Unveiling the Two "Superpowers" Behind AI Video Creationoriginal ↗
2 May 2025
立委科普：揭秘AI创作视频的两种“神功”original ↗
1 May 2025
非量化自回归视频生成模型NOVA的技术路线original ↗
1 May 2025
生成式AI的两条视频生成路线original ↗
30 Apr 2025
MCP: From Flashy Boom to Real Usability — A Technical Deep Diveoriginal ↗
26 Apr 2025
MCP：从“伪繁荣”到可落地的进化路线original ↗
26 Apr 2025
Silicon Valley Night: A Foxy Encounteroriginal ↗
23 Apr 2025
《硅谷夜记：艳遇》original ↗
22 Apr 2025
短剧：黄石的低语 (Whispers of Yellowstone)original ↗
3 Apr 2025
breaking news：黄石起义：熊的宣言original ↗
3 Apr 2025
自传体小说《刀锋人生：百年缝合》（3/3）original ↗
3 Apr 2025
自传体小说《刀锋人生：百年缝合》（2）original ↗
3 Apr 2025
自传体小说：《刀锋人生：百年缝合》original ↗
2 Apr 2025
The Scalpel’s Edge: A Life Stitched Through a Century (3）original ↗
2 Apr 2025
The Scalpel’s Edge: A Life Stitched Through a Century (2）original ↗
2 Apr 2025
The Scalpel’s Edge: A Life Stitched Through a Century (自传体小说）original ↗
2 Apr 2025
CHAPTER 15: RECENT GATHERING SPEECHESoriginal ↗
2 Apr 2025
CHAPTER 16: THE LI FAMILY VALUESoriginal ↗
2 Apr 2025
CHAPTER 7: SEASONS OF WIND AND RAIN (1)original ↗
2 Apr 2025
CHAPTER 5: SEASONS OF WIND AND RAINoriginal ↗
2 Apr 2025
CHAPTER 4: THE BURNING SUNSET GLOWoriginal ↗
2 Apr 2025
CHAPTER 3: MY SURGICAL CAREER—OUTSIDE THE HOSPITALoriginal ↗
2 Apr 2025
https://liweinlp.com/13134original ↗
2 Apr 2025
CHAPTER 11: CHRONICLES OF LIFE-AND-DEATH EXPERIENCESoriginal ↗
2 Apr 2025
CHAPTER 10: MY LIFE WITH THE 127TH HOSPITALoriginal ↗
2 Apr 2025
CHAPTER 13: ANOTHER COLLEGE ENTRANCE EXAMINATION APPROACHESoriginal ↗
2 Apr 2025
CHAPTER 14: SWEET – TANIA'S BRILLIANT LIFEoriginal ↗
2 Apr 2025
中文分词的前世今生original ↗
29 Mar 2025
大模型如何解锁AI各种任务成为通用引擎的？original ↗
28 Mar 2025
Grok：大模型为什么要超大数据？（4o配图）original ↗
27 Mar 2025
Grok: 大力出奇迹的背后original ↗
27 Mar 2025
Grok：超大数据的大模型为何能收敛？original ↗
26 Mar 2025
《“蜜蜂巢”里的子弹：JFK档案解密后》original ↗
24 Mar 2025
2025年肯尼迪遇刺档案解密揭示的惊人真相original ↗
24 Mar 2025
2025年肯尼迪遇刺案档案解密original ↗
24 Mar 2025
Gemini Deep Research：用“Logits Lens”洞察神经网络的奥秘original ↗
22 Mar 2025
03 deep research： Challenges and Prospects of Advanced Reasoning LLMsoriginal ↗
20 Mar 2025
Sonnet3.7: 推理大模型的挑战与前景（图文版）original ↗
20 Mar 2025
Gemini deep research podcast： Challenges and Prospects of Advanced Reasoning LLMsoriginal ↗
20 Mar 2025
数学圆舞曲：欧拉恒等式（配乐诗朗诵）original ↗
19 Mar 2025
检索增强（RAG）与窗口数据的互补性 (图文版）original ↗
19 Mar 2025
检索增强（RAG）与窗口数据的互补性original ↗
19 Mar 2025
人类 vs 恐龙：一场关于“不作不死”的滑稽短剧original ↗
16 Mar 2025
deep research: 最新颈椎病手术指征与治疗概览original ↗
15 Mar 2025
关于颈椎病，大模型医疗建议靠谱吗？original ↗
14 Mar 2025
给奶奶讲一下AI最新物种“大模型代理”original ↗
12 Mar 2025
Decoding LLM-native Agents: Bridging Compilation and Interpretation in AIoriginal ↗
12 Mar 2025
Xiao Hong (Red): The Man Behind the Autonomus Genral Agent Manusoriginal ↗
10 Mar 2025
The Agent Era: The Contemporary Evolution from Chatbots to Digital Agentsoriginal ↗
10 Mar 2025
o3 deep research: 智能体的分野和演进original ↗
10 Mar 2025
肖弘与Manus：AI Agent 的实战方法论original ↗
10 Mar 2025
Xiao Hong Red：肖弘其人original ↗
9 Mar 2025
万字长文解析 LLM-native Agent 及其混合计算方式original ↗
9 Mar 2025
o3 deep research: LLM 驱动的 Agent 综述original ↗
7 Mar 2025
Agent：数字代理的崛起与未来original ↗
7 Mar 2025
Agent元年：从聊天机器人到数字员工的当代进化史original ↗
7 Mar 2025
【外一篇：推理范式演进中的概念】original ↗
6 Mar 2025
生成式AI学习中容易混淆的几个术语original ↗
5 Mar 2025
思维链是大模型的符号神助攻original ↗
4 Mar 2025
再谈自然模态数据是高维空间的低维流形original ↗
4 Mar 2025
The Three-Stage Scaling Laws Large Language Modelsoriginal ↗
3 Mar 2025
大模型三阶段的 scaling laws 接力赛original ↗
3 Mar 2025
Fundamental Limitations of Deep Learning: Origins in Data-Driven Methodologyoriginal ↗
2 Mar 2025
深度学习的局限性研究综述original ↗
2 Mar 2025
o3 deep research: 深度学习局限性研究报告original ↗
2 Mar 2025
深度学习的基石：多层感知机original ↗
28 Feb 2025
左脚踩右脚可以飞吗，谈交替使用监督微调和强化学习的后训练original ↗
27 Feb 2025
o3 Deep Research: DeepSeek R1 多阶段训练流程问答解析original ↗
27 Feb 2025
Dilemma of RPA and Early-Stage LLM Co-pilot Entrepreneurs in the Age of Agent Tsunamioriginal ↗
27 Feb 2025
RPA 赛道与大模型Copilots早期创业者的困局original ↗
26 Feb 2025
Understanding the Division of Labor Among Q, K, V in Self-Attention Mechanismoriginal ↗
25 Feb 2025
Linguists Should Find Self-Attention Intuitively Familiaroriginal ↗
24 Feb 2025
语言学家应该很容易理解自注意力机制original ↗
24 Feb 2025
Introduction to Transformer and Its Attention Mechanismoriginal ↗
23 Feb 2025
Has Symbolism Been Sidelined for Too Long? Could Neural LLM terminate AI?original ↗
23 Feb 2025
符号主义被打入冷宫太久了，难道神经是AI的终结者吗？original ↗
23 Feb 2025
Transformer 和注意力机制简介original ↗
20 Feb 2025
立委科普：如何理解自注意力机制中的QKV分工？original ↗
20 Feb 2025
True Story Behind DeepSeek's Success: AI Learning to Think Slowly Without Human Supervisionoriginal ↗
15 Feb 2025
DeepSeek爆火真相：不靠“人盯”， 让AI自己学会慢思考original ↗
15 Feb 2025
Does the New Reasoning Paradigm (Query+CoT+Answer) Support a New Scaling Law?original ↗
14 Feb 2025
Technical Deep Dive: Understanding DeepSeek R1's Reasoning Mechanism in Productionoriginal ↗
14 Feb 2025
DeepSeek 笔记：推理新范式 query+cot+answer 支持新的 scaling law 吗？original ↗
13 Feb 2025
DeepSeek 笔记：R1 部署阶段的推理机制original ↗
13 Feb 2025
Hallucinations in AI: Bug or Feature? A Deep Dive into DeepSeek-R1original ↗
9 Feb 2025
从R1幻觉谈起，大模型幻觉是缺陷还是创意火花？original ↗
8 Feb 2025
Previous Next