7 papers
Bio-inspired Learning and Decision-Making with Probabilistic In-Memory Computing Hardware: Part 2
Thomas Dalgaty, Eiji Kawasaki, Miguel de Prado +3
This report extends our previous work (Part 1), which introduced an energy-based model for learning and decision-making under uncertainty. The model leverages stochastic Langevin d…
Bio-inspired Learning and Decision-Making with Probabilistic In-Memory Computing Hardware: Part 1
Thomas Dalgaty, Eiji Kawasaki, Miguel de Prado +2
Learning and decision-making in animals are often modeled as Bayesian processes, where sensory evidence is integrated with prior beliefs to guide behavior in the face of uncertaint…
Contrastive Regularization of Machine Learning Potentials
Dimitrios Tzivrailis, Georgios Sotiropoulos, Alberto Rosso +1
Machine learning models are increasingly trained as regressors and then deployed as generators, from language models to learned simulators. In this setting, accuracy on held-out da…
AI estimation of partition functions with the PULSE method for the thermodynamic properties of disordered compounds
Baptiste Bernard, Luca Messina, Eiji Kawasaki +1
Predicting the thermodynamic properties of chemically disordered materials is constrained by the combinatorial growth of possible atomic configurations, which makes conventional sa…
Uncertainty in AI-driven Monte Carlo simulations
Dimitrios Tzivrailis, Alberto Rosso, Eiji Kawasaki
In the study of complex systems, evaluating physical observables often requires sampling representative configurations via Monte Carlo techniques. These methods rely on repeated ev…
Targeting the partition function of chemically disordered materials with a generative approach based on inverse variational autoencoders
Maciej J. Karcz, Luca Messina, Eiji Kawasaki +1
Computing atomic-scale properties of chemically disordered materials requires an efficient exploration of their vast configuration space. Traditional approaches such as Monte Carlo…
Scaling-up Memristor Monte Carlo with magnetic domain-wall physics
Thomas Dalgaty, Shogo Yamada, Anca Molnos +8
By exploiting the intrinsic random nature of nanoscale devices, Memristor Monte Carlo (MMC) is a promising enabler of edge learning systems. However, due to multiple algorithmic an…