Predicting cellular responses to perturbations requires interpretable insights into molecular regulatory dynamics to perform reliable cell fate control, despite the confounding non-linearity of the underlying interactions. There is a growing interest in developing machine learning-based perturbation response prediction models to handle the non-linearity of perturbation data, but their interpretation in terms of molecular regulatory dynamics remains a challenge. Alternatively, for meaningful biological interpretation, logical network models such as Boolean networks are widely used in systems biology to represent intracellular molecular regulation. However, determining the appropriate regulatory logic of large-scale networks remains an obstacle due to the high-dimensional and discontinuous search space. To tackle these challenges, we present a scalable derivative-free optimizer trained by meta-reinforcement learning for Boolean network models. The logical network model optimized by the trained optimizer successfully predicts anti-cancer drug responses of cancer cell lines, while simultaneously providing insight into their underlying molecular regulatory mechanisms. Display omitted •We propose a gray box framework with Boolean networks and a black box optimizer•GREY is a learned optimizer specialized in Boolean network optimization•Gray box framework can predict drug responses and reveal underlying mechanisms The challenge of predicting cellular responses to perturbations amid the complex non-linearities of molecular interactions has spurred the development of machine learning-based models. However, interpreting these models in terms of molecular regulatory dynamics remains difficult. Conversely, logical network models like Boolean networks offer interpretability but struggle with large-scale networks due to high-dimensional search spaces. To overcome these hurdles, we introduce a scalable derivative-free optimizer, trained via meta-reinforcement learning, for Boolean network models. This approach enables prediction of anti-cancer drug responses in cancer cell lines while offering valuable insights into their molecular regulatory mechanisms. Kim et al. present a gray box framework that combines a white box logical model with a black box optimizer, addressing challenges in interpreting molecular regulatory dynamics. The gray box framework successfully predicts anti-cancer drug responses of cancer cells, while shedding light on the underlying molecular regulatory mechanisms.