{"ID":2860984,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.03051","arxiv_id":"2510.03051","title":"ZeroShotOpt: Towards Zero-Shot Pretrained Models for Efficient Black-Box Optimization","abstract":"Global optimization of expensive, derivative-free black-box functions requires extreme sample efficiency. While Bayesian optimization (BO) is the current state-of-the-art, its performance hinges on surrogate and acquisition function hyper-parameters that are often hand-tuned and fail to generalize across problem landscapes. We present ZeroShotOpt, a general-purpose, pretrained model for continuous black-box optimization tasks ranging from 2D to 20D. Our approach leverages offline reinforcement learning on large-scale optimization trajectories collected from 12 BO variants. To scale pretraining, we generate millions of synthetic Gaussian process-based functions with diverse landscapes, enabling the model to learn transferable optimization policies. As a result, ZeroShotOpt achieves robust zero-shot generalization on a wide array of unseen benchmarks, matching or surpassing the sample efficiency of leading global optimizers, including BO, while also offering a reusable foundation for future extensions and improvements. Our open-source code, dataset, and model are available at: https://github.com/jamisonmeindl/zeroshotopt","short_abstract":"Global optimization of expensive, derivative-free black-box functions requires extreme sample efficiency. While Bayesian optimization (BO) is the current state-of-the-art, its performance hinges on surrogate and acquisition function hyper-parameters that are often hand-tuned and fail to generalize across problem landsc...","url_abs":"https://arxiv.org/abs/2510.03051","url_pdf":"https://arxiv.org/pdf/2510.03051v1","authors":"[\"Jamison Meindl\",\"Yunsheng Tian\",\"Tony Cui\",\"Veronika Thost\",\"Zhang-Wei Hong\",\"Johannes Dürholt\",\"Jie Chen\",\"Wojciech Matusik\",\"Mina Konaković Luković\"]","published":"2025-10-03T14:33:23Z","proceeding":"cs.LG","tasks":"[\"cs.LG\",\"cs.AI\"]","methods":"[\"Reinforcement Learning\"]","has_code":false,"code_links":[{"ID":608781,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_id":2860984,"paper_url":"https://arxiv.org/abs/2510.03051","paper_title":"ZeroShotOpt: Towards Zero-Shot Pretrained Models for Efficient Black-Box Optimization","repo_url":"https://github.com/jamisonmeindl/zeroshotopt","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}