https://eigenroom.com/blogSangho An's personal notes on robot learning, papers, and building things.enhttps://eigenroom.com/blog/project-ruleshttps://eigenroom.com/blog/project-rulesSat, 26 Sep 2026 00:00:00 GMTPut 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.https://eigenroom.com/blog/groot-n1https://eigenroom.com/blog/groot-n1Mon, 21 Sep 2026 00:00:00 GMTA reading of GR00T N1, from its two-system architecture to the data pipeline that connects different robot embodiments.https://eigenroom.com/blog/planning-before-codehttps://eigenroom.com/blog/planning-before-codeMon, 14 Sep 2026 00:00:00 GMTThe 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.https://eigenroom.com/blog/diffusion-policy-part-2https://eigenroom.com/blog/diffusion-policy-part-2Mon, 10 Aug 2026 00:00:00 GMTSections 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.https://eigenroom.com/blog/diffusion-policy-part-1https://eigenroom.com/blog/diffusion-policy-part-1Mon, 03 Aug 2026 00:00:00 GMTSections 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.https://eigenroom.com/blog/understanding-diffusion-modelshttps://eigenroom.com/blog/understanding-diffusion-modelsWed, 10 Jun 2026 00:00:00 GMTFollowing Calvin Luo's unified perspective in order: the ELBO, hierarchical VAEs, variational diffusion models, the three equivalent prediction targets, and the score-based view.https://eigenroom.com/blog/vae-and-elbohttps://eigenroom.com/blog/vae-and-elboFri, 01 May 2026 00:00:00 GMTWhy 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.https://eigenroom.com/blog/autoencodershttps://eigenroom.com/blog/autoencodersSun, 26 Apr 2026 00:00:00 GMTHow 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.https://eigenroom.com/blog/probability-notation-guidehttps://eigenroom.com/blog/probability-notation-guideWed, 22 Apr 2026 00:00:00 GMTp(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.https://eigenroom.com/blog/mle-cross-entropy-klhttps://eigenroom.com/blog/mle-cross-entropy-klTue, 21 Apr 2026 00:00:00 GMTMaximum 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.https://eigenroom.com/blog/probability-for-generative-modelshttps://eigenroom.com/blog/probability-for-generative-modelsTue, 21 Apr 2026 00:00:00 GMTWhy a deep-learning series has to start with probability: what distributions, Bayes' theorem, expectation, and Jensen's inequality each contribute to generative models.https://eigenroom.com/blog/diffusion-from-zero-introhttps://eigenroom.com/blog/diffusion-from-zero-introMon, 06 Apr 2026 00:00:00 GMTWhy 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.