@article{
vu2026onpolicy,
title={On-Policy Model Error Suffices:  An Invariant-Measure Return-Gap Bound for Model-Based Reinforcement Learning},
author={Minh Nhat VU},
journal={Transactions on Machine Learning Research},
issn={2835-8856},
year={2026},
url={https://openreview.net/forum?id=qDJ8axLhMa},
note={}
}