Home

Blog Posts

Technical articles on machine learning, computational chemistry, and scientific computing with deep dives into methodologies and practical implementations.
Generative Modeling
MNIST digit samples generated from a Variational Autoencoder latent space

Importance Weighted Autoencoders: Beyond the Standard VAE

The key difference between multi-sample VAEs and IWAEs: how log-of-averages creates a tighter bound on log-likelihood.

Computational Chemistry
The transformation from a 2D chemical structure image to a SMILES representation

What is Optical Chemical Structure Recognition (OCSR)?

A micro-review of Optical Chemical Structure Recognition (OCSR), from rule-based systems to modern AI models.

Computational Chemistry
SELFIES strings guarantee 100% valid molecules - even when generated randomly

Converting SELFIES Strings to 2D Molecular Images

Visualize SELFIES molecular representations and test their 100% robustness through random sampling experiments.

Computational Chemistry
Aspirin molecular structure generated from SMILES string

Converting SMILES Strings to 2D Molecular Images

Learn how to create 2D molecular images from SMILES strings using RDKit and PIL, with proper formatting and legends.

Scientific Computing
Comparison of exponential sampling methods showing histograms from both inverse transform and von Neumann methods overlaid with the theoretical exponential distribution

Exponential Random Number Generation: Two Classic Algorithms Compared

Compare inverse transform sampling and von Neumann's rejection method for exponential random numbers with Python …

Computational Chemistry
Müller-Brown Potential Energy Surface showing the three minima and two saddle points

Implementing the Müller-Brown Potential in PyTorch

Guide to implementing the Müller-Brown potential in PyTorch, comparing analytical vs automatic differentiation with …

Scientific Computing
Velocity Autocorrelation Function showing the signature negative region characteristic of liquid dynamics

Digital Restoration: Modernizing Rahman's 1964 Argon Simulation

How I used modern software engineering (caching, vectorization, and dependency locking) to reproduce a 60-year-old …...

Computational Chemistry
Comparison of 2D molecular graph versus 3D conformer ensemble showing latanoprost molecule in multiple conformations

GEOM Dataset: 3D Molecular Conformer Generation

Learn how GEOM transforms 2D molecular graphs into dynamic 3D conformer ensembles for molecular machine learning …

Generative Modeling
Variational Autoencoder architecture diagram showing encoder, latent space, and decoder

Modern PyTorch Techniques for VAEs: A Hands-On Tutorial

Learn to implement VAEs in PyTorch: ELBO objective, reparameterization trick, loss scaling, and MNIST experiments on …

Computational Chemistry
Comparison chart showing k-NN significantly outperforming logistic regression for molecular classification across different alkane sizes

Can You Hear the Shape of a Molecule? (Part Three)

Supervised learning reveals hidden eigenvalue patterns that clustering missed, testing k-NN and logistic regression on …

Computational Chemistry
Charts showing Dunn Index, distance metrics, and computation time analysis revealing clustering performance degradation with molecular size

Can You Hear the Shape of a Molecule? (Part Two)

Clustering analysis reveals why Coulomb matrix eigenvalues struggle with larger alkanes, using Dunn Index and silhouette …

Natural Language Processing
Word vector illustration showing text classification and NLP concepts

Sarcasm Detection with Transformers: A Cautionary Tale

Learn how dataset bias can lead to misleading results in NLP: a sarcasm detection model that actually learned to …