Benchmarking Protein Ensemble Generators

Generative models for protein conformational ensembles are arriving monthly, but the field has no neutral cross-model comparison. A tour of why ensemble evaluation is hard and where the methods landscape sits today.

June 2, 2026 · 7 min · Jacob Silterra

From One Algorithm to Five: Rebuilding Parametric t-SNE as Parametric DR

Rebuilding parametric_tsne from scratch as parametric_dr: a PyTorch library providing parametric neural network implementations of five dimensionality reduction methods under a single API.

March 11, 2026 · 10 min · Jacob Silterra