Policy Adaptation

Policy adaptation focuses on enabling artificial agents, such as robots or software systems, to quickly and effectively adjust their behavior to new, unseen situations or environments. Current research emphasizes efficient methods for adapting existing policies, leveraging techniques like transformer-based encoders, hypernetworks, and optimal transport for policy fusion, often in conjunction with reinforcement learning algorithms. This research is significant because it addresses the critical limitations of traditional methods that struggle with generalization and robustness, paving the way for more adaptable and reliable AI systems in diverse real-world applications, including robotics, healthcare, and resource management.

Papers

November 18, 2024

Continual Task Learning through Adaptive Policy Self-Composition
Shengchao Hu, Yuhang Zhou, Ziqing Fan, Jifeng Hu, Li Shen, Ya Zhang, Dacheng Tao
Catastrophic Forgetting Multiple Agent Continual Learning Benchmark Generalist Agent Policy Adaptation

October 22, 2024

Survival of the Fittest: Evolutionary Adaptation of Policies for Environmental Shifts
Sheryl Paul, Jyotirmoy V. Deshmukh
Reinforcement Learning Robust Policy Prior Policy Policy Adaptation Deep Survival Environmental Change

October 16, 2024

Robust RL with LLM-Driven Data Synthesis and Policy Adaptation for Autonomous Driving
Sihao Wu, Jiaxu Liu, Xiangyu Yin, Guangliang Cheng, Meng Fang, Xingyu Zhao, Xinping Yi, Xiaowei Huang
Reinforcement Learning Autonomous Driving LLM Agent Robust Reinforcement Learning Data Synthesis Policy Adaptation

October 13, 2024

Meta-Reinforcement Learning with Universal Policy Adaptation: Provable Near-Optimality under All-task Optimum Comparator
Siyuan Xu, Minghui Zhu
Reinforcement Learning Meta Reinforcement Learning Near Optimality Policy Adaptation Policy Network

September 26, 2024

LoopSR: Looping Sim-and-Real for Lifelong Policy Adaptation of Legged Robots
Peilin Wu, Weiji Xie, Jiahang Cao, Hang Lai, Weinan Zhang
Reinforcement Learning Real World Legged Robot Legged Locomotion Policy Adaptation Continual Training Policy Generation

August 29, 2024

Policy Adaptation via Language Optimization: Decomposing Tasks for Few-Shot Imitation
Vivek Myers, Bill Chunyuan Zheng, Oier Mees, Sergey Levine, Kuan Fang
Vision Language Model Feature Imitation New Task Language Conditioned Shot Adaptation Policy Adaptation

July 10, 2024

Examining Long-Context Large Language Models for Environmental Review Document Comprehension
Hung Phan, Anurag Acharya, Rounak Meyur, Sarthak Chaturvedi, Shivam Sharma, Mike Parker, Dan Nally, Ali Jannesari, Karl Pazdernik, Mahantesh Halappanavar, Sai Munikoti, Sameera Horawalavithana
Language Model Long Context Long Context Large Language Model Question Answering System Policy Adaptation Kg Rag

May 31, 2024

Fusion-PSRO: Nash Policy Fusion for Policy Space Response Oracles
Jiesong Lian
Nash Equilibrium Policy Adaptation Policy Synthesis Policy Space Response Oracle Transitive Game

April 1, 2024

CAAP: Class-Dependent Automatic Data Augmentation Based On Adaptive Policies For Time Series
Tien-Yu Chang, Hao Dai, Vincent S. Tseng
Data Augmentation Time Series Class Imbalance Augmentation Method Policy Adaptation Adaptive Policy

March 20, 2024

Federated reinforcement learning for robot motion planning with zero-shot generalization
Zhenyuan Yuan, Siyuan Xu, Minghui Zhu
Zero Shot Federated Reinforcement Learning Robot Motion Planning Policy Adaptation Zero Shot Generalization

February 29, 2024

Go Beyond Black-box Policies: Rethinking the Design of Learning Agent for Interpretable and Verifiable HVAC Control
Zhiyu An, Xianzhong Ding, Wan Du
Product Design Model Based Reinforcement Learning Black Box Model Interpretable Way Learning Agent Policy Adaptation Heating, Ventilation, and Air Conditioning

February 12, 2024

SemTra: A Semantic Skill Translator for Cross-Domain Zero-Shot Policy Adaptation
Sangwoo Shin, Minjong Yoo, Jeongwoo Lee, Honguk Woo
Semantic Understanding Policy Adaptation Skill Adaptation

February 9, 2024

Distilling Morphology-Conditioned Hypernetworks for Efficient Universal Morphology Control
Zheng Xiong, Risto Vuorio, Jacob Beck, Matthieu Zimmer, Kun Shao, Shimon Whiteson
Zero Shot Policy Adaptation Policy Distillation Reinforcement Learning Policy Conditioned Hypernetwork Multi Morphology Controller

January 29, 2024

Zero-shot Imitation Policy via Search in Demonstration Dataset
Federco Malato, Florian Leopold, Andrew Melnik, Ville Hautamaki
Zero Shot Latent Representation Search Query Behavior Cloning Policy Adaptation Demonstration Data Minecraft Agent

November 13, 2023

Combinatorial Optimization with Policy Adaptation using Latent Space Search
Felix Chalumeau, Shikha Surana, Clement Bonnet, Nathan Grinsztajn, Arnu Pretorius, Alexandre Laterre, Thomas D. Barrett
Reinforcement Learning Policy Adaptation Capacitated Vehicle Routing Problem

October 8, 2023

Safe Deep Policy Adaptation
Wenli Xiao, Tairan He, John Dolan, Guanya Shi
Reinforcement Learning Autonomous Robot Safe Reinforcement Learning Policy Adaptation Adaptive Policy

September 30, 2023

AdaptNet: Policy Adaptation for Physics-Based Character Control
Pei Xu, Kaixiang Xie, Sheldon Andrews, Paul G. Kry, Michael Neff, Morgan McGuire, Ioannis Karamouzas, Victor Zordan
Reinforcement Learning Physic Based Adaptive Network Policy Adaptation Policy Network

April 5, 2023

Convex Optimization-based Policy Adaptation to Compensate for Distributional Shifts
Navid Hashemi, Justin Ruths, Jyotirmoy V. Deshmukh
Distribution Shift Optimal Trajectory Adaptive Cruise Control Optimal Controller Policy Adaptation Real World Trajectory

January 16, 2023

Neuro-Symbolic World Models for Adapting to Open World Novelty
Jonathan Balloch, Zhiyu Lin, Robert Wright, Xiangyu Peng, Mustafa Hussain, Aarun Srinivas, Julia Kim, Mark O. Riedl
Novelty Search Policy Adaptation New Environment

December 14, 2022

Policy Adaptation from Foundation Model Feedback
Yuying Ge, Annabella Macaluso, Li Erran Li, Ping Luo, Xiaolong Wang
Foundation Model Vision Language Foundation Model Policy Adaptation Conditioned Policy Policy Fine Tuning

Policy Adaptation

Papers

Continual Task Learning through Adaptive Policy Self-Composition

Survival of the Fittest: Evolutionary Adaptation of Policies for Environmental Shifts

Robust RL with LLM-Driven Data Synthesis and Policy Adaptation for Autonomous Driving

Meta-Reinforcement Learning with Universal Policy Adaptation: Provable Near-Optimality under All-task Optimum Comparator

LoopSR: Looping Sim-and-Real for Lifelong Policy Adaptation of Legged Robots

Policy Adaptation via Language Optimization: Decomposing Tasks for Few-Shot Imitation

Examining Long-Context Large Language Models for Environmental Review Document Comprehension

Fusion-PSRO: Nash Policy Fusion for Policy Space Response Oracles

CAAP: Class-Dependent Automatic Data Augmentation Based On Adaptive Policies For Time Series

Federated reinforcement learning for robot motion planning with zero-shot generalization

Go Beyond Black-box Policies: Rethinking the Design of Learning Agent for Interpretable and Verifiable HVAC Control

SemTra: A Semantic Skill Translator for Cross-Domain Zero-Shot Policy Adaptation

Distilling Morphology-Conditioned Hypernetworks for Efficient Universal Morphology Control

Zero-shot Imitation Policy via Search in Demonstration Dataset

Combinatorial Optimization with Policy Adaptation using Latent Space Search

Safe Deep Policy Adaptation

AdaptNet: Policy Adaptation for Physics-Based Character Control

Convex Optimization-based Policy Adaptation to Compensate for Distributional Shifts

Neuro-Symbolic World Models for Adapting to Open World Novelty

Policy Adaptation from Foundation Model Feedback