The field of optimization is moving towards more efficient and autonomous methods, with a focus on multi-objective optimization and machine learning. Recent developments have introduced new algorithms and frameworks that can efficiently solve complex optimization problems, such as Global-Order GFlowNets and Contrastive Learning-based Constraint Reordering. These methods have shown promise in various applications, including robotics, materials science, and smart manufacturing. Another trend is the use of GPU acceleration to improve the computational efficiency of optimization algorithms, such as in the case of TensorNSGA-III. Additionally, there is a growing interest in autonomous experimentation and Bayesian optimization, which can guide the search for optimal solutions and provide physical insights into the trade-offs between different objectives. Notable papers include:

• Global-Order GFlowNets, which resolves conflicts in optimization objectives by transforming local order into global order.
• CLCR, which accelerates MILP solving by optimizing constraint ordering using contrastive learning.
• MOUR-QD, which extends Quality-Diversity algorithms to unstructured and unbounded feature spaces, achieving state-of-the-art results in robotic tasks.