Stop re-explaining your project: writing an AGENTS.md that works
Put your project's standing decisions in one file your coding agent reads every time, and write it yourself: research suggests AI-generated rule files don't help.
FIELD NOTES BY SANGHO AN
Learning,
one note
at a time.
Robot learning, papers read slowly, and the things I figure out by building.
Reading GR00T N1: connecting the islands of robot data
A reading of GR00T N1, from its two-system architecture to the data pipeline that connects different robot embodiments.
A reading note
The notebook
Reading GR00T N1: connecting the islands of robot data
A reading of GR00T N1, from its two-system architecture to the data pipeline that connects different robot embodiments.
Prompt, plan, verify: how I hand coding work to an AI
The hard part of AI-assisted coding has moved from writing code to checking it. This is the routine I use to keep that checking small: structure the request, review the plan, then test the result myself.
Reading Diffusion Policy, part 2: why it trains stably, and what the robots show
Sections 4 to 9 of Diffusion Policy: modes and basins, why learning the score removes the normalizing constant, the benchmark and real-robot results, and the limits the authors state.
Reading Diffusion Policy, part 1: from formulation to network design
Sections 1 to 3 of Diffusion Policy: why behavior cloning needs multimodal action distributions, how DDPM becomes a visuomotor policy, and the CNN, transformer, and FiLM design choices.
Reading Understanding Diffusion Models: from VAEs to score-based models
Following Calvin Luo's unified perspective in order: the ELBO, hierarchical VAEs, variational diffusion models, the three equivalent prediction targets, and the score-based view.
VAEs and the ELBO: learning a distribution instead of a point
Why an autoencoder isn't enough for generation, how a VAE learns a latent distribution, what each term of the ELBO does, and how the reparameterization trick makes it trainable.
Autoencoders: compression, bottlenecks, and why they aren't quite generative
How an autoencoder compresses and reconstructs its input, why the bottleneck matters, what reconstruction loss means in probability terms, and why a plain autoencoder is hard to sample from.
What each probability actually means: a field guide to VAE and diffusion notation
p(x), p(z), p(x|z), p(z|x), q, and p-theta: what each expression in VAEs and diffusion models describes, explained with concrete examples instead of derivations.
Where MSE and cross-entropy come from: maximum likelihood and KL divergence
Maximum likelihood explains the losses we use every day: a Gaussian assumption gives MSE, a categorical one gives cross-entropy, and both connect to KL divergence.
Distributions, Bayes' theorem, expectation, and Jensen's inequality: the four ideas behind the ELBO
Why a deep-learning series has to start with probability: what distributions, Bayes' theorem, expectation, and Jensen's inequality each contribute to generative models.
Diffusion from Zero: learning generative models from the ground up
Why I started this series as an electrical engineering student with no probability course behind me, how the parts fit together, and the resources that helped most.