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
Transformer
Published:
In this blog, I will introduce the transformer architecture, which is the foundation of modern large language models. For better illustration, I will take translation task as an example.
Embedding
Published:
In the world of machine learning, we use embeddings(mathematically speaking, vectors) to represent everything. In this post, I will introduce how to encode nature languages into embeddings.
Attention
Published:
In this blog, we will focus on the attention mechanism, which is a key component of the Transformer architecture.
Flow Matching and Diffusion Models 3
Published:
In our previous post, we defined the theoretical targets for Flow and Diffusion models. However, calculating these global targets directly is often mathematically intractable.In this post, we bridge the gap between theory and practice. We will explore:
Flow Matching and Diffusion Models 2
Published:
In our previous post, we explored how Flow and Diffusion models define trajectories using ODEs and SDEs. However, a key question remains: What exactly should the neural network learn to approximate?
Flow Matching and Diffusion Models 1
Published:
This post serves as the first entry in a series introcuding Flow Matching and Diffusion Models, potentially spanning a broad range of related topics as I expand upon areas of personal and research interest.
Special credits to MIT.
publications
