Sitemap
A list of all the posts and pages found on the site. For you robots out there, there is an XML version available for digesting as well.
Pages
Posts
深度生成模型（2）—— 变分自编码器 VAE
Published:
本文介绍变分自编码器（Variational Autoencoder, VAE）的原理和实现。
参考课程：Stanford CS236，MIT 6s978
1 VAE的原理
1.1 第一步
现在有一个MNIST手写数据集，数据集包含很多的图片，每一张图片的大小为 ，像素的灰度有十个等级 ，并且每张…
👍 知乎赞同：5
深度生成模型（1）—— 自回归模型 AR
Published:
本文介绍生成模型的基础——自回归模型（Autoregressive Model）。
参考课程：Stanford CS236，MIT 6s978
1 从一切的开始说起
1.1 生成模型是什么？
生成无处不在，我们说的每一句话，都是我们的大脑生成的，我们看到的每一张图片，也是现实世界生成给我们的…
👍 知乎赞同：6
深度生成模型（7）—— 扩散模型 Diffusion
Published:
本文深入介绍扩散模型（Diffusion Model）的原理，包括 DDPM 等经典方法。
参考资料：Stanford CS236，MIT 6s978，Song Yang’s Blog，DDPM论文，向diffusion更深处：score matching、随机微分方程和朗之万动力学
1 DDPM
…
👍 知乎赞同：7
深度生成模型（3）—— 流模型 Normalizing Flow Model
Published:
本文介绍归一化流模型（Normalizing Flow）的原理及其在生成模型中的应用。
参考课程：Stanford CS236
1 Flow Model 的定义
1.1 密度变换公式 Change of Variables Theorem
上期我们谈到，VAE 使用变分下界 ELBO 绕过了…
👍 知乎赞同：9
深度生成模型（6）—— 分数模型 Score-Based Model
Published:
本文介绍分数模型（Score-Based Model）的原理及其与能量模型的联系。
参考课程：Stanford CS236，Song Yang’s Blog
0 从 Energy-based Model 到 Score-based Model
上期我们讲到在能量模型中，可以用分数匹配代替…
👍 知乎赞同：10
一文讲明白LM中的RL（PPO，RLHF，DPO，GRPO，……）
Published:
本文系统梳理了语言模型中的强化学习方法，包括 PPO、RLHF、DPO、GRPO 等核心算法。
0 前置知识
RL从何而来：强化学习（1）—— 从模仿学习说起
策略梯度：强化学习（2）—— 策略梯度法（REINFORCE）
Actor-Critic：强化学习（3）—— Actor-Critic
Q-Le…
👍 知乎赞同：12
深度生成模型（5）—— 能量模型 Energy-Based Model (EBM)
Published:
本文介绍能量模型（Energy-Based Model, EBM）的基本概念和原理。
参考课程：Stanford CS236
1 EBM的定义
1.1 其他模型的问题
在第二节中我们讲到，VAE 通过多个正态分布叠加对目标分布进行学习，对于MNIST或许足够，但是对于尺寸更大、彩色的图片来说…
👍 知乎赞同：28
portfolio
Portfolio item number 1
Short description of portfolio item number 1
Portfolio item number 2
Short description of portfolio item number 2
publications
Paper Title Number 1
Published in Journal 1, 2009
This paper is about the number 1. The number 2 is left for future work.
Recommended citation: Your Name, You. (2009). "Paper Title Number 1." Journal 1. 1(1).
Download Paper | Download Slides
Paper Title Number 2
Published in Journal 1, 2010
This paper is about the number 2. The number 3 is left for future work.
Recommended citation: Your Name, You. (2010). "Paper Title Number 2." Journal 1. 1(2).
Download Paper | Download Slides
Paper Title Number 3
Published in Journal 1, 2015
This paper is about the number 3. The number 4 is left for future work.
Recommended citation: Your Name, You. (2015). "Paper Title Number 3." Journal 1. 1(3).
Download Paper | Download Slides
Paper Title Number 4
Published in GitHub Journal of Bugs, 2024
This paper is about fixing template issue #693.
Recommended citation: Your Name, You. (2024). "Paper Title Number 3." GitHub Journal of Bugs. 1(3).
Download Paper
resources
Mathematical Analysis 1 (数学分析 1)
Published:
This paper is about the number 1. The number 2 is left for future work.
Mathematical Analysis 2 (数学分析 2)
Published:
This paper is about the number 2. The number 3 is left for future work.
Advanced Algebra 1 (高等代数 1)
Published:
This paper is about the number 3. The number 4 is left for future work.
Advanced Algebra 2 (高等代数 2)
Published:
This paper is about fixing template issue #693.
talks
Talk 1 on Relevant Topic in Your Field
Published:
This is a description of your talk, which is a markdown file that can be all markdown-ified like any other post. Yay markdown!
Conference Proceeding talk 3 on Relevant Topic in Your Field
Published:
This is a description of your conference proceedings talk, note the different field in type. You can put anything in this field.
teaching
Teaching experience 1
Undergraduate course, University 1, Department, 2014
This is a description of a teaching experience. You can use markdown like any other post.
Teaching experience 2
Workshop, University 1, Department, 2015
This is a description of a teaching experience. You can use markdown like any other post.