Why Multi-Task Learning Reduces Accuracy: The Mathematical Inevitability of Simultaneous Optimization for Multiple Objectives

If foundation models like CNNs and BERT are strong at multiple tasks, then training them on multiple tasks simultaneously ...
US Geological Survey

Managing and learning with multiple models: Objectives and optimization algorithms

Authors William J. M. Probert, C.E. Hauser, E. McDonald-Madden, M.C. Runge, P.W.J. Baxter, H.P. Possingham ...