Thermodynamic Integration

Thermodynamic integration (TI) is a computational technique used to estimate free energy differences, a crucial quantity in various scientific fields. Current research focuses on applying TI principles to diverse areas, including machine learning model development (e.g., using neural networks to improve efficiency and accuracy of TI calculations), materials science (simulating microstructure evolution and predicting thermodynamic properties), and even the analysis of language models and human cognition. This approach offers a powerful framework for improving the accuracy and efficiency of complex simulations and predictions across disciplines, leading to advancements in areas such as materials design, drug discovery, and the understanding of complex systems.

Papers

July 26, 2022

Thermodynamics of learning physical phenomena
Elias Cueto, Francisco Chinesta
Inductive Bias Thermodynamic Integration Learning Process

May 18, 2022

Counting Phases and Faces Using Bayesian Thermodynamic Integration
Alexander Lobashev, Mikhail V. Tamm
Human Face Posterior Distribution Thermodynamic Integration Non Equilibrium Fisher Information Metric Quantitative Phase

May 1, 2022

Thermodynamically Consistent Machine-Learned Internal State Variable Approach for Data-Driven Modeling of Path-Dependent Materials
Xiaolong He, Jiun-Shyan Chen
Data Driven Data Driven Modeling Thermodynamic Integration Path Dependent Stress Strain

April 10, 2022

Model-free optimization of power/efficiency tradeoffs in quantum thermal machines using reinforcement learning
Paolo Andrea Erdman, Frank Noé
Reinforcement Learning Quantum Physic Trade Offs Quantum Machine Thermal Management Thermodynamic Integration Non Equilibrium Positive Friction Model Free Optimization

December 31, 2021

Separation of Scales and a Thermodynamic Description of Feature Learning in Some CNNs
Inbar Seroussi, Gadi Naveh, Zohar Ringel
Convolutional Neural Network Deep Neural Network Deep Convolutional Neural Network Feature Learning Separation Performance Neural Network Output Thermodynamic Integration Ghanaian Musical Scale Faster Model Infinite Width Neural Network

November 4, 2021

Variational Inference with Holder Bounds
Junya Chen, Danni Lu, Zidi Xiu, Ke Bai, Lawrence Carin, Chenyang Tao
Variational Inference Variational Method General Bound Variational Objective Thermodynamic Integration