VRVishwaksen Reddy Dhareddyinvishwaksen.hashnode.dev·Jun 1 · 6 min readDoes Hashing Make Text Look Random?This post is the first part of a small experiment series on statistical structure in text, hashes, and encrypted data. The question is simple: if I take a list of English words and hash each word, doe00
VRVishwaksen Reddy Dhareddyinvishwaksen.hashnode.dev·Jun 1 · 6 min readWhy Unigram KL Is Not EnoughThis is Part 3 of the series on KL divergence, hashing, and encryption outputs. The first two posts used byte frequencies as the main lens. That worked for plaintext, hashes, and some weak ciphers, bu00
VRVishwaksen Reddy Dhareddyinvishwaksen.hashnode.dev·Jun 1 · 6 min readAncient Ciphers Hide Text, But Do They Destroy Structure?This is Part 2 of the KL divergence experiment series. Part 1 showed that hashing a wordlist makes the byte distribution look much closer to uniform. This post asks a messier question: what happens wh00
ВМВлад Мешковinbeta-divergence.hashnode.dev·Dec 6, 2025 · 8 min readOutlier-Robust Variational Inference: Making Deep Learning More ResilientTable of Contents Introduction A Quick Refresher on Bayesian Inference The Variational Inference Idea KL Divergence and the ELBO Where ELBO Shows Up in Practice Why Standard Variational Inference Fails with Outliers Robust Divergences: β and γ...00
KNKunal Nayyarinkunal221.hashnode.dev·Jun 5, 2025 · 17 min readKL DivergenceRemember when we chatted about how AI models like Generative Adversarial Networks (GANs) learn to create? Well, there's another class of AI magicians : Variational Autoencoders (VAEs) and Diffusion Models, that achieve incredible feats of generation ...00
