Dynamic Generalization

Dynamic generalization in machine learning focuses on developing models capable of extrapolating learned behaviors to unseen dynamic systems, even with varying parameters or environmental conditions. Current research emphasizes incorporating contextual information into model architectures, such as through neural ordinary differential equations (NODEs) with context flows or adaptive context-aware policies, often leveraging meta-learning techniques and modular designs to improve generalization performance. This research is crucial for advancing reinforcement learning, enabling more robust and adaptable AI agents in complex, real-world scenarios, and facilitating the development of more generalizable scientific models for diverse physical systems.

Papers

May 3, 2024

Neural Context Flows for Meta-Learning of Dynamical Systems
Roussel Desmond Nzoyem, David A.W. Barton, Tom Deakin
Neural Ordinary Differential Equation Dynamic Generalization

October 25, 2023

Dynamics Generalisation in Reinforcement Learning via Adaptive Context-Aware Policies
Michael Beukman, Devon Jarvis, Richard Klein, Steven James, Benjamin Rosman
Reinforcement Learning Strong Generalization Transition Dynamic Behaviour Learning Dynamic Generalization

January 3, 2023

Metalearning generalizable dynamics from trajectories
Qiaofeng Li, Tianyi Wang, Vwani Roychowdhury, M. Khalid Jawed
Dynamical System Neural Ordinary Differential Equation Handwritten Trajectory Meta Training Metalearning Algorithm Dynamic Generalization

November 23, 2022

Prototypical context-aware dynamics generalization for high-dimensional model-based reinforcement learning
Junjie Wang, Yao Mu, Dong Li, Qichao Zhang, Dongbin Zhao, Yuzheng Zhuang, Ping Luo, Bin Wang, Jianye Hao
High Dimensional World Model Dynamic Generalization Context Generalization

October 9, 2022

Decomposed Mutual Information Optimization for Generalized Context in Meta-Reinforcement Learning
Yao Mu, Yuzheng Zhuang, Fei Ni, Bin Wang, Jianyu Chen, Jianye Hao, Ping Luo
Context Learning Mutual Information Meta Reinforcement Learning Transition Dynamic Dynamic Generalization

June 28, 2022

Optimal Estimation of Generic Dynamics by Path-Dependent Neural Jump ODEs
Florian Krach, Marc Nübel, Josef Teichmann
Neural ODE Optimal Estimation Dynamic Generalization General Stochastic

June 24, 2022

ModLaNets: Learning Generalisable Dynamics via Modularity and Physical Inductive Bias
Yupu Lu, Shijie Lin, Guanqi Chen, Jia Pan
Physic Informed Neural Network Modularity Function Lagrangian System Lagrangian Neural Network Dynamic Generalization

Dynamic Generalization

Papers

Neural Context Flows for Meta-Learning of Dynamical Systems

Dynamics Generalisation in Reinforcement Learning via Adaptive Context-Aware Policies

Metalearning generalizable dynamics from trajectories

Prototypical context-aware dynamics generalization for high-dimensional model-based reinforcement learning

Decomposed Mutual Information Optimization for Generalized Context in Meta-Reinforcement Learning

Optimal Estimation of Generic Dynamics by Path-Dependent Neural Jump ODEs

ModLaNets: Learning Generalisable Dynamics via Modularity and Physical Inductive Bias