<?xml version="1.0" encoding="utf-8"?><feed xmlns="http://www.w3.org/2005/Atom" ><generator uri="https://jekyllrb.com/" version="4.3.3">Jekyll</generator><link href="https://nathanjzhao.github.io/blog/feed/tagged/ml.xml" rel="self" type="application/atom+xml" /><link href="https://nathanjzhao.github.io/" rel="alternate" type="text/html" /><updated>2026-03-04T09:31:59+00:00</updated><id>https://nathanjzhao.github.io/blog/feed/tagged/ml.xml</id><title type="html">Nathan’s Notes</title><subtitle>A blog about things.</subtitle><entry><title type="html">Psycholinguistics and Transformer Circuits</title><link href="https://nathanjzhao.github.io/psycholinguistics" rel="alternate" type="text/html" title="Psycholinguistics and Transformer Circuits" /><published>2025-10-15T00:00:00+00:00</published><updated>2025-10-15T00:00:00+00:00</updated><id>https://nathanjzhao.github.io/psycholinguistics</id><author><name></name></author><category term="ml" /><category term="musings" /><summary type="html"><![CDATA[Exploring the intricate web of word recognition in the brain, in relation to language models]]></summary></entry><entry><title type="html">Discretizing Continuous SSMs</title><link href="https://nathanjzhao.github.io/blelloch" rel="alternate" type="text/html" title="Discretizing Continuous SSMs" /><published>2025-09-23T00:00:00+00:00</published><updated>2025-09-23T00:00:00+00:00</updated><id>https://nathanjzhao.github.io/blelloch</id><author><name></name></author><category term="ml" /><category term="math" /><summary type="html"><![CDATA[The math behind why RNN-ifying SSMs works.]]></summary></entry><entry><title type="html">MRI Reconstruction and Convex Optimization</title><link href="https://nathanjzhao.github.io/mri-reconstruction" rel="alternate" type="text/html" title="MRI Reconstruction and Convex Optimization" /><published>2024-08-02T00:00:00+00:00</published><updated>2024-08-02T00:00:00+00:00</updated><id>https://nathanjzhao.github.io/mri-reconstruction</id><author><name></name></author><category term="ml" /><summary type="html"><![CDATA[An accessible introduction to MRI Reconstruction and solving them with unrolled neural networks]]></summary></entry><entry><title type="html">Mamba: Selective State Space Modeling</title><link href="https://nathanjzhao.github.io/mamba" rel="alternate" type="text/html" title="Mamba: Selective State Space Modeling" /><published>2024-07-22T00:00:00+00:00</published><updated>2024-07-22T00:00:00+00:00</updated><id>https://nathanjzhao.github.io/mamba</id><author><name></name></author><category term="ml" /><category term="math" /><summary type="html"><![CDATA[An introduction to Mamba models: faster and better* than transformers]]></summary></entry><entry><title type="html">Flow Matching</title><link href="https://nathanjzhao.github.io/flow-matching" rel="alternate" type="text/html" title="Flow Matching" /><published>2024-06-17T00:00:00+00:00</published><updated>2024-06-17T00:00:00+00:00</updated><id>https://nathanjzhao.github.io/flow-matching</id><author><name></name></author><category term="ml" /><category term="math" /><summary type="html"><![CDATA[A overview of diffusion models and flow matching models.]]></summary></entry></feed>