Constrained generative modeling is fundamental to applications such as robotic control and autonomous driving, where models must respect physical laws and safety-critical constraints. In real-world settings, these constraints rarely take the form of simple linear inequalities, but instead complex feasible regions that resemble road maps or other structured spatial domains. We propose a constrained generation framework that generates samples directly within such feasible regions while preserving realism. Our method fine-tunes a pretrained generative model to enforce constraints while maintaining generative fidelity. Experimentally, our method exhibits characteristics distinct from existing fine-tuning and training-free constrained baselines, revealing a new compromise between constraint satisfaction and sampling quality.
@article{arxiv.2603.06742,
title = {Improved Constrained Generation by Bridging Pretrained Generative Models},
author = {Xiaoxuan Liang and Saeid Naderiparizi and Yunpeng Liu and Berend Zwartsenberg and Frank Wood},
journal= {arXiv preprint arXiv:2603.06742},
year = {2026}
}