analiza aiqbench
This position paper challenges the industry-standard "train-then-fix" paradigm by arguing that real-world agents must inherently support continual learning due to inevitable post-deployment non-stationarity. The authors identify four specific sources of environmental drift that necessitate never-ending adaptation, suggesting that optimal deployed systems are those that continuously refine their policies rather than waiting for performance degradation. This shift is critical for practitioners aiming to build robust RL systems that maintain efficacy in dynamic environments without costly periodic retraining.
Position: Deployed Reinforcement Learning should be Continual
arXiv:2606.04029v1 Announce Type: new Abstract: Reinforcement Learning (RL) has received increasing attention and adoption in real-world use cases. Most of these systems follow a train-then-fix paradigm, where trained agents do not learn while interacting with the world until performance degrades and retraining becomes necessary. In this position paper, we argue that deploying an agent that is incapable of optimality, but receives an evaluative reward signal, is inherently a continual RL problem. We identify four sources of non-stationarity after deployment that necessitate never-ending learning, and highlight why the best deployed agents never stop adapting. We analyze successful examples of continual RL in the real world, and present the community with the advantages and measures to move away from the current train-then-fix paradigm.
Abstract:Reinforcement Learning (RL) has received increasing attention and adoption in real-world use cases. Most of these systems follow a train-then-fix paradigm, where trained agents do not learn while interacting with the world until performance degrades and retraining becomes necessary. In this position paper, we argue that deploying an agent that is incapable of optimality, but receives an evaluative reward signal, is inherently a continual RL problem. We identify four sources of non-stationarity after deployment that necessitate never-ending learning, and highlight why the best deployed agents never stop adapting. We analyze successful examples of continual RL in the real world, and present the community with the advantages and measures to move away from the current train-then-fix paradigm.
Accepted to the ICML 2026 Position Paper Track. See this https URL
Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
arXiv:2606.04029 [cs.LG]
(or arXiv:2606.04029v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2606.04029 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 223 cuvinte
Subiecte conexe
CIDER: Continual Interactive Distillation for Embodied Reinforcement Learning
arXiv · 25 aug. 2026
Beyond Single-Model Optimization: Preserving Plasticity in Continual Reinforcement Learning
arXiv · 10 iun. 2026
NeuroSynth: A Biologically Inspired Continual Reinforcement Learning Architecture for Mitigating Catastrophic Forgetting
arXiv · 3 aug. 2026
Continual Graph Multi-Agent Reinforcement Learning
arXiv · 8 oct. 2026