Generative-Models

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

Summary of Burda, Grosse & Salakhutdinov's ICLR 2016 paper introducing Importance Weighted Autoencoders for tighter …

Major updates to the SELFIES library, improved performance, expanded chemistry support, and new customization features.

The 2020 paper introducing SELFIES, the 100% robust molecular representation that solves SMILES validity problems in ML …

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

SELFIES is a 100% robust molecular string representation for ML, implemented in the open-source selfies Python library.

Liu et al.'s ICLR 2025 paper introducing DenoiseVAE, which learns adaptive, atom-specific noise for better molecular …

Aneja et al.'s NeurIPS 2021 paper introducing Noise Contrastive Priors (NCPs) to address VAE's 'prior hole' problem with …

Skinnider (2024) shows that generating invalid SMILES actually improves chemical language model performance through …

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

Perspective on SELFIES as a 100% robust SMILES alternative, with 16 future research directions for molecular AI.

Complete guide to GAN objective functions including WGAN, LSGAN, Fisher GAN, and more. Understand which loss function to …