Instance Generation for Meta-Black-Box Optimization Through Latent Space Reverse Engineering

Chen Wang; Yue-Jiao Gong; Zhiguang Cao; Zeyuan Ma

AAAI 2026technical0 citations

Instance Generation for Meta-Black-Box Optimization Through Latent Space Reverse Engineering

Chen Wang, Yue-Jiao Gong, Zhiguang Cao, Zeyuan Ma

Abstract

To relieve intensive human-expertise required to design optimization algorithms, recent Meta-Black-Box Optimization (MetaBBO) researches leverage generalization strength of meta-learning to train neural network-based algorithm design policies over a predefined training problem set, which automates the adaptability of the low-level optimizers on unseen problem instances. Currently, a common training problem set choice in existing MetaBBOs is well-known benchmark suites CoCo-BBOB. Although such choice facilitates the MetaBBO

BibTeX

@inproceedings{aaai2026_instancegenerati,
  title = {Instance Generation for Meta-Black-Box Optimization Through Latent Space Reverse Engineering},
  author = {Chen Wang and Yue-Jiao Gong and Zhiguang Cao and Zeyuan Ma},
  booktitle = {AAAI 2026},
  year = {2026}
}