paper-with-me

Papers

Online Reinforcement Learning in Non-Stationary Context-Driven Environments

2023-02-04 · Pouya Hamadanian, Arash Nasr-Esfahany, Malte Schwarzkopf, Siddartha Sen, Mohammad Alizadeh

We study online reinforcement learning (RL) in non-stationary environments, where a time-varying exogenous context process affects the environment dynamics. Online RL is challenging in such environments due to "catastrophic forgetting" (CF). The agent tends to forget prior knowledge as it trains on new experiences. Prior approaches to mitigate this issue assume task labels (which are often not available in practice), employ brittle regularization heuristics, or use off-policy methods that suffer from instability and poor performance. We present Locally Constrained Policy Optimization (LCPO), an online RL approach that combats CF by anchoring policy outputs on old experiences while optimizing the return on current experiences. To perform this anchoring, LCPO locally constrains policy optimization using samples from experiences that lie outside of the current context distribution. We evaluate LCPO in Mujoco, classic control and computer systems environments with a variety of synthetic and real context traces, and find that it outperforms a variety of baselines in the non-stationary setting, while achieving results on-par with a "prescient" agent trained offline across all context traces. LCPO's source code is available at https://github.com/pouyahmdn/LCPO.

📄 PDF Abstract BibTeX arXiv:2302.02182

Code (2)

lcpo-rl/lcpo 공식 구현 pytorch
pouyahmdn/lcpo 공식 구현 pytorch

Tasks

MuJoCoreinforcement-learningReinforcement LearningReinforcement Learning (RL)

Similar Papers 제목 키워드 기반

Safe Continual Reinforcement Learning Methods for Nonstationary Environments. Towards a Survey of the State of the Art

2026-01-08 · Timofey Tomashevskiy arxiv

This work provides a state-of-the-art survey of continual safe online reinforcement learning (COSRL) methods. We discuss theoretical aspects, challenges, and open questions in building continual online safe reinforcement…

Reinforcement LearningContinual Learning

A Robust Policy Bootstrapping Algorithm for Multi-objective Reinforcement Learning in Non-stationary Environments

2023-08-18 · Sherif Abdelfattah, Kathryn Kasmarik, Jiankun Hu

Multi-objective Markov decision processes are a special kind of multi-objective optimization problem that involves sequential decision making while satisfying the Markov property of stochastic processes. Multi-objective …

Decision MakingMulti-Objective Reinforcement Learningreinforcement-learningReinforcement Learning+1

Optimal Dynamic Regret by Transformers for Non-Stationary Reinforcement Learning

2025-08-22 · Baiyuan Chen, Shinji Ito, Masaaki Imaizumi arxiv

Transformers have demonstrated exceptional performance across a wide range of domains. While their ability to perform reinforcement learning in-context has been established both theoretically and empirically, their behav…

Reinforcement Learning

Context-Aware Safe Reinforcement Learning for Non-Stationary Environments

2021-01-02 · Baiming Chen, Zuxin Liu, Jiacheng Zhu, Mengdi Xu 외

Safety is a critical concern when deploying reinforcement learning agents for realistic tasks. Recently, safe reinforcement learning algorithms have been developed to optimize the agent's performance while avoiding viola…

Meta-Learningreinforcement-learningReinforcement LearningReinforcement Learning (RL)+1

Online Adaptive Reinforcement Learning with Echo State Networks for Non-Stationary Dynamics

2026-02-06 · Aoi Yoshimura, Gouhei Tanaka arxiv

Reinforcement learning (RL) policies trained in simulation often suffer from severe performance degradation when deployed in real-world environments due to non-stationary dynamics. While Domain Randomization (DR) and met…

Computational EfficiencyReinforcement Learning