Removing nodal and support-mismatch pathologies in Variational Monte Carlo via blurred sampling
The episode discusses a paper using blurred sampling to remove nodal and support-mismatch pathologies in Variational Monte Carlo (VMC). The hosts explain that this post-processing technique stabilizes stochastic estimators, leading to finite-sample unbiased results. This method preserves computational scaling while resolving infinite variance caused by nodes and support mismatches, making VMC more reliable for practical quantum simulations.