no code implementations • 23 Jan 2023 • Elisabeth Agoritsas, Giovanni Catania, Aurélien Decelle, Beatriz Seoane
In this paper, we quantify the impact of using non-convergent Markov chains to train Energy-Based models (EBMs).
no code implementations • 21 Jan 2021 • Bastien Marguet, Elisabeth Agoritsas, Léonie Canet, Vivien Lecomte
We show that, contrarily to the common belief, it is possible to extend the known reversible construction to the case of arbitrary irreversible dynamics, for overdamped Langevin equations with additive white noise - provided their steady state is known.
Statistical Mechanics Mathematical Physics Mathematical Physics