发表机构
Saarland University; Universidade Estadual de Campinas（萨尔兰大学; 坎皮纳斯州立大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
MOSEL框架通过将后验多目标优化转化为Stackelberg博弈，以单次反向传播实现帕累托驻留前沿的高效学习，在公平性-准确性等冲突场景中表现优异，并提升多任务学习效果。
AI 中文摘要
我们提出了MOSEL（多目标Stackelberg高效学习）框架，用于深度神经网络中的后验多目标优化（MOO），该框架以标准单目标训练的计算成本恢复完整的帕累托驻留解前沿。MOSEL将问题重新表述为双层优化问题，利用网络模块化将表示学习与目标偏好对齐解耦。将双层优化问题视为Stackelberg博弈，使得能够在单次前向-后向传播中解决原始的后验多目标优化问题。因此，MOSEL在匹配标准单目标训练的时间和内存效率的同时，实现了可扩展的帕累托驻留前沿学习。实验上，MOSEL在强冲突场景（如公平性-准确性）中发现了多样且最优的帕累托前沿。值得注意的是，即使在弱冲突场景（如多任务学习）中，它也持续收敛到更接近乌托邦点的解，优于标准单目标训练和专门的多任务学习方法。这些结果凸显了后验多目标优化学习作为高效学习更多样、更稳健表示路径的更广泛潜力，最终提升泛化能力。
英文摘要
We propose MOSEL (Multi-Objective Stackelberg Efficient Learning), a framework for a posteriori multi-objective optimization (MOO) in deep neural networks that recovers a full front of Pareto stationary solutions at the computational cost of standard single-objective training. MOSEL reformulates the problem as a bilevel optimization problem that leverages network modularity to decouple representation learning from objective-preference alignment. Casting the bilevel optimization problem as a Stackelberg game enables solving the original a posteriori MOO problem in a single forward-backward pass. As a result, MOSEL matches the time and memory efficiency of standard single-objective training while enabling scalable Pareto stationary front learning. Empirically, MOSEL uncovers diverse and optimal Pareto frontiers in strongly conflicting settings (e.g., fairness-accuracy). Remarkably, even in weakly conflicting regimes such as multi-task learning, it consistently converges to solutions closer to the utopia point, outperforming both standard single-objective training and specialized multi-task learning methods. These results highlight the broader potential of a posteriori MOO learning as a pathway to efficiently learn more diverse and robust representations, ultimately improving generalization.
