The spelled-out intro to neural networks and backpropagation: building micrograd
17 人推荐 · 跨 16 个月 · 2022–2025
中文导读
Andrej Karpathy 经典神课《神经网络：从零到英雄》第一讲。仅凭基础 Python 和高中微积分，纯手写构建出微型自动求导引擎 micrograd，手把手推导反向传播与神经网络训练全流程。
为什么 HN 推它
彻底搞懂反向传播与深度学习底层机理的首选教程。Karpathy 不讲空泛概念，而是用代码一行行带你建立直觉、理解算法设计的“为什么”。
「这门课含金量极高，因为他不仅讲是什么和怎么做，更从直觉出发透彻解释了背后的底层原因。」HN 网友原话：It's especially useful because he explains the why and provides intuitive explanations instead of just talking about the what and how.
英文原摘要（HackerTube / Claude 整理）
Andrej Karpathy's first Neural Networks: Zero to Hero lecture, which builds micrograd, a tiny autograd engine, from scratch in Python. It explains backpropagation and neural network training step by step, assuming only basic Python and some high-school calculus.
HN treats it as the best starting point for really understanding backprop and neural nets, because Karpathy explains the why with intuition and builds everything in code.
HN 上的提及
Karpathy has a relatively small youtube channel he started a few weeks ago.
I recommend watching Andrej Karpathy's recent videos on backprop and neural nets. Makes it really simple and intuitive.
first video is literally from scratch (i.e. just Python
Andrej has a brilliant series of videos on youtube where he explores the basics of neural nets right up to transformer architecture through code.