arXiv Machine Learning June 25, 2026 By Philipp Grohs, Davide Nobile

Is Variational Monte Carlo Robust? Sharp Moment Thresholds and Heavy-tailed Stochastic Optimization

Read the original on arXiv Machine Learning →

arXiv:2606. 26009v1 Announce Type: new Abstract: Variational Monte Carlo (VMC) is a central algorithm in electronic structure theory and has gained renewed importance through modern neural-network ans\"atze such as FermiNet.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.

arXiv AI

Oct 26

Implicit Neural Representations for Variational Problems on Graphons

arXiv:2610. 32147v1 Announce Type: cross Abstract: We propose a neural exploratory framework for variational problems over graphons, i.

By Taeyoung Kim, Jineon Baek, Joonkyung Lee, Hongseok Yang

arXiv Machine Learning

Jun 15

Direct/adaptive-mixture phase-gradient learning for neural-network quantum states with complex phase structure

arXiv:2606. 13912v1 Announce Type: cross Abstract: Neural-network quantum states (NQS) are a leading variational tool for quantum many-body physics, yet their optimization is fragile whenever the ground state carries a non-trivial sign or complex phase structure, a situation generic to gauge fields, broken time-reversal symmetry, and fermionic statistics.

By Yi-Ran Xue, Rui Wang, Baigeng Wang, Chenan Wei

arXiv Machine Learning

Sep 14

Almost Sure Convergence Analysis of Stochastic Gradient Methods with Clipping and Additive Noise

The paper proves that stochastic gradient descent with gradient clipping and additive Gaussian noise (SGD‑CN) converges almost surely under smoothness and bounded noise assumptions, given standard decaying step sizes. The analysis extends to momentum variants such as the stochastic heavy ball and Nesterov's accelerated gradient, showing that careful energy constructions yield similar guarantees. These results provide stronger theoretical foundations for understanding the pathwise behaviour of clipped stochastic gradient methods in both convex and nonconvex regimes.

By Amartya Mukherjee, Jun Liu

safety

arXiv Machine Learning

Sep 30

High-Probability Convergence of Clipped SGD under Heavy-Tailed Noise and $(L_0,L_1)$-Smoothness

arXiv:2505.20817v3 Announce Type: replace-cross Abstract: Gradient clipping is widely used in language-model training to control heavy-tailed gradient noise and can improve convergence guarantees ove...

By Taha El Bakkali El Kadi, Savelii Chezhegov, Aleksandr Beznosikov, Samuel Horv\'ath, Eduard Gorbunov

safety

arXiv Statistics ML

Sep 30

A stochastic subgradient method with optimal failure exponent

arXiv:2609.37425v1 Announce Type: cross Abstract: Fix a target accuracy $\varepsilon$, a gradient-noise level $s$, and a horizon $N$. We wish to design algorithms which minimize the probability of ob...

By Bart P. G. van Parys

safety

arXiv Machine Learning

Sep 22

Gradient-estimator design overcomes trainability barriers in neural-network-based variational optimization

arXiv:2609.22342v1 Announce Type: cross Abstract: Neural networks provide expressive representations for scientific computing. However, even sufficiently expressive networks can suffer training failu...

By Yi-Ran Xue, Rui Wang, Baigeng Wang, Chenan Wei

fine-tuning

One email a morning, machine-written

Related stories

Implicit Neural Representations for Variational Problems on Graphons

Direct/adaptive-mixture phase-gradient learning for neural-network quantum states with complex phase structure

Almost Sure Convergence Analysis of Stochastic Gradient Methods with Clipping and Additive Noise

High-Probability Convergence of Clipped SGD under Heavy-Tailed Noise and $(L_0,L_1)$-Smoothness

A stochastic subgradient method with optimal failure exponent

Gradient-estimator design overcomes trainability barriers in neural-network-based variational optimization