Reproduce “Stochastic Latent Twins: Learning Stochastic Solution Operators” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Matthias Chung, Jack Michael Solomon, Deepanshu Verma
Stochastic Latent Twins learn conditional distributions between physical states at chosen times, avoiding recursive small-step propagation at inference. A deterministic encoder and decoder surround a probabilistic latent transition, where uncertainty resides; compatible constraints preserve invariants. In the tested periodic Burgers experiment, every physics-conforming prediction preserved mass and had lower median and sliced-Wasserstein errors than Latent NSF, with mild later underdispersion.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.