Meta Reinforcement Learning

Meta-reinforcement learning (Meta-RL) aims to create agents capable of rapidly adapting to new tasks with minimal experience, leveraging prior learning to accelerate adaptation. Current research focuses on improving sample efficiency and generalization across diverse tasks, employing model architectures like recurrent neural networks, transformers, and hypernetworks, along with algorithms such as MAML and PPO, often incorporating techniques like importance sampling and constrained optimization. This field is significant for its potential to enable more robust and adaptable AI systems in various applications, from resource optimization in wireless networks to safe and efficient robot control in unpredictable environments.

Papers

October 6, 2022

Meta Reinforcement Learning for Optimal Design of Legged Robots
Álvaro Belmonte-Baeza, Joonho Lee, Giorgio Valsecchi, Marco Hutter
Legged Robot Quadrupedal Robot Meta Reinforcement Learning Design Optimization Optimal Design Robot Design

October 1, 2022

Zero-Shot Policy Transfer with Disentangled Task Representation of Meta-Reinforcement Learning
Zheng Wu, Yichen Xie, Wenzhao Lian, Changhao Wang, Yanjiang Guo, Jianyu Chen, Stefan Schaal, Masayoshi Tomizuka
Reinforcement Learning Zero Shot Meta Reinforcement Learning Compositional Task Policy Generalization

September 26, 2022

Enhanced Meta Reinforcement Learning using Demonstrations in Sparse Reward Environments
Desik Rengarajan, Sapana Chaudhary, Jaewon Kim, Dileep Kalathil, Srinivas Shakkottai
Sparse Reward Noisy Demonstration Meta Reinforcement Learning Sparse Reward Environment Robust Meta Reinforcement Learning

September 21, 2022

September 19, 2022

Meta-Reinforcement Learning for Adaptive Control of Second Order Systems
Daniel G. McClement, Nathan P. Lawrence, Michael G. Forbes, Philip D. Loewen, Johan U. Backström, R. Bhushan Gopaluni
Meta Learning Second Order Meta Reinforcement Learning Adaptive Control Model Free Policy Generalizable Controller

September 13, 2022

Designing Biological Sequences via Meta-Reinforcement Learning and Bayesian Optimization
Leo Feng, Padideh Nouri, Aneri Muni, Yoshua Bengio, Pierre-Luc Bacon
Bayesian Optimization Meta Reinforcement Learning Biological Sequence Autoregressive Generative Model Sequence Design Likely Sequence

September 11, 2022

Meta-Reinforcement Learning via Language Instructions
Zhenshan Bing, Alexander Koch, Xiangtong Yao, Kai Huang, Alois Knoll
Reinforcement Learning Deep Reinforcement Learning Natural Language Instruction Meta Reinforcement Learning Meta Rl Algorithm Language Instruction Manipulation Benchmark

August 24, 2022

A model-based approach to meta-Reinforcement Learning: Transformers and tree search
Brieuc Pinon, Jean-Charles Delvenne, Raphaël Jungers
Transformer Megatron Decepticons Meta Learning Model Based Model Free Meta Reinforcement Learning Tree Search Meta Rl Algorithm

July 29, 2022

July 19, 2022

Learning Action Translator for Meta Reinforcement Learning on Sparse-Reward Tasks
Yijie Guo, Qiucheng Wu, Honglak Lee
Sparse Reward Meta Reinforcement Learning Meta Training Policy Adaptation

July 13, 2022

Continual Meta-Reinforcement Learning for UAV-Aided Vehicular Wireless Networks
Riccardo Marini, Sangwoo Park, Osvaldo Simeone, Chiara Buratti
Reinforcement Learning Meta Reinforcement Learning Vehicular Network Base Station

June 21, 2022

Meta Reinforcement Learning with Finite Training Tasks -- a Density Estimation Approach
Zohar Rimon, Aviv Tamar, Gilad Adler
Density Estimation Meta Reinforcement Learning Task Distribution Meta Rl Algorithm Bayes Optimal Policy

June 14, 2022

Transformers are Meta-Reinforcement Learners
Luckeciano C. Melo
Transformer Megatron Decepticons Transformer Architecture Meta Reinforcement Learning

June 7, 2022

May 31, 2022

A Meta Reinforcement Learning Approach for Predictive Autoscaling in the Cloud
Siqiao Xue, Chao Qu, Xiaoming Shi, Cong Liao, Shiyi Zhu, Xiaoyu Tan, Lintao Ma, Shiyu Wang, Shijun Wang, Yun Hu, Lei Lei, Yangfei Zheng, Jianguo Li, James Zhang
Meta Reinforcement Learning Cloud Computing Optimal Scaling Auto Scaling Workload Forecasting

May 25, 2022

Fast Inference and Transfer of Compositional Task Structures for Few-shot Task Generalization
Sungryull Sohn, Hyunjae Woo, Jongwook Choi, lyubing qiang, Izzeddin Gur, Aleksandra Faust, Honglak Lee
Formality Transfer Hierarchical Reinforcement Learning Meta Reinforcement Learning Fast Inference Task Graph Task Inference Shot Generalization Task Structure

April 29, 2022

Line of Sight Curvature for Missile Guidance using Reinforcement Meta-Learning
Brian Gaudet, Roberto Furfaro
Meta Reinforcement Learning Best Fit Line Agile Maneuver Attitude Estimation Guidance Trajectory Dynamic Target

Meta Reinforcement Learning

Papers

Meta Reinforcement Learning for Optimal Design of Legged Robots

Zero-Shot Policy Transfer with Disentangled Task Representation of Meta-Reinforcement Learning

Enhanced Meta Reinforcement Learning using Demonstrations in Sparse Reward Environments

Learning from Symmetry: Meta-Reinforcement Learning with Symmetrical Behaviors and Language Instructions

On the Convergence Theory of Meta Reinforcement Learning with Personalized Policies

Meta-Reinforcement Learning for Adaptive Control of Second Order Systems

Designing Biological Sequences via Meta-Reinforcement Learning and Bayesian Optimization

Meta-Reinforcement Learning via Language Instructions

A model-based approach to meta-Reinforcement Learning: Transformers and tree search

Sampling Attacks on Meta Reinforcement Learning: A Minimax Formulation and Complexity Analysis

Meta Reinforcement Learning with Successor Feature Based Context

Learning Action Translator for Meta Reinforcement Learning on Sparse-Reward Tasks

Continual Meta-Reinforcement Learning for UAV-Aided Vehicular Wireless Networks

Meta Reinforcement Learning with Finite Training Tasks -- a Density Estimation Approach

Transformers are Meta-Reinforcement Learners

On the Effectiveness of Fine-tuning Versus Meta-reinforcement Learning

Variational Meta Reinforcement Learning for Social Robotics

A Meta Reinforcement Learning Approach for Predictive Autoscaling in the Cloud

Fast Inference and Transfer of Compositional Task Structures for Few-shot Task Generalization

Line of Sight Curvature for Missile Guidance using Reinforcement Meta-Learning