https://nesl42.github.io/tags/llm/ 敬渊's Blog Hugo 0.154.5 & FixIt v0.4.3-20260123080729-2a5bd268 zh-CN Sun, 27 Jul 2025 00:00:00 +0000 https://nesl42.github.io/posts/202507-alphagopaperread/ Sun, 27 Jul 2025 00:00:00 +0000 https://nesl42.github.io/posts/202507-alphagopaperread/ Technology <h1 class="heading-element" id="论文阅读笔记模型架构发现的alphago时刻"><span><strong>论文阅读笔记：模型架构发现的AlphaGo时刻</strong></span> <a href="#%e8%ae%ba%e6%96%87%e9%98%85%e8%af%bb%e7%ac%94%e8%ae%b0%e6%a8%a1%e5%9e%8b%e6%9e%b6%e6%9e%84%e5%8f%91%e7%8e%b0%e7%9a%84alphago%e6%97%b6%e5%88%bb" class="heading-mark"> <svg class="octicon octicon-link" viewBox="0 0 16 16" version="1.1" width="16" height="16" aria-hidden="true"><path d="m7.775 3.275 1.25-1.25a3.5 3.5 0 1 1 4.95 4.95l-2.5 2.5a3.5 3.5 0 0 1-4.95 0 .751.751 0 0 1 .018-1.042.751.751 0 0 1 1.042-.018 1.998 1.998 0 0 0 2.83 0l2.5-2.5a2.002 2.002 0 0 0-2.83-2.83l-1.25 1.25a.751.751 0 0 1-1.042-.018.751.751 0 0 1-.018-1.042Zm-4.69 9.64a1.998 1.998 0 0 0 2.83 0l1.25-1.25a.751.751 0 0 1 1.042.018.751.751 0 0 1 .018 1.042l-1.25 1.25a3.5 3.5 0 1 1-4.95-4.95l2.5-2.5a3.5 3.5 0 0 1 4.95 0 .751.751 0 0 1-.018 1.042.751.751 0 0 1-1.042.018 1.998 1.998 0 0 0-2.83 0l-2.5 2.5a1.998 1.998 0 0 0 0 2.83Z"></path></svg> </a> </h1><h2 class="heading-element" id="摘要"><span><strong>摘要</strong></span> <a href="#%e6%91%98%e8%a6%81" class="heading-mark"> <svg class="octicon octicon-link" viewBox="0 0 16 16" version="1.1" width="16" height="16" aria-hidden="true"><path d="m7.775 3.275 1.25-1.25a3.5 3.5 0 1 1 4.95 4.95l-2.5 2.5a3.5 3.5 0 0 1-4.95 0 .751.751 0 0 1 .018-1.042.751.751 0 0 1 1.042-.018 1.998 1.998 0 0 0 2.83 0l2.5-2.5a2.002 2.002 0 0 0-2.83-2.83l-1.25 1.25a.751.751 0 0 1-1.042-.018.751.751 0 0 1-.018-1.042Zm-4.69 9.64a1.998 1.998 0 0 0 2.83 0l1.25-1.25a.751.751 0 0 1 1.042.018.751.751 0 0 1 .018 1.042l-1.25 1.25a3.5 3.5 0 1 1-4.95-4.95l2.5-2.5a3.5 3.5 0 0 1 4.95 0 .751.751 0 0 1-.018 1.042.751.751 0 0 1-1.042.018 1.998 1.998 0 0 0-2.83 0l-2.5 2.5a1.998 1.998 0 0 0 0 2.83Z"></path></svg> </a> </h2><p>尽管人工智能系统（AI systems）展现出指数级提升的能力，但人工智能研究（AI research）本身的速度却受限于人类的认知能力（human cognitive capacity），这造成了日益严重的开发瓶颈（development bottleneck）。本文介绍了 ASI-ARCH，这是人工智能研究领域（ASI4AI）中人工超级智能（Artificial Superintelligence for AI research）的首次演示，它在神经架构发现（neural architecture discovery）这一关键领域，通过使人工智能能够进行自身的架构创新（architectural innovation），打破了这一根本限制。</p> https://nesl42.github.io/posts/202503-mcp/ Wed, 19 Mar 2025 00:00:00 +0000 https://nesl42.github.io/posts/202503-mcp/ Technology <p>最近 MCP 很火啊，但是生态还不是很完善，等等看几个月之后的 MCP 是什么情况吧。</p>