January 31, 2025
2025
A new preprint Deep Multi-Task Learning Has Low Amortized Intrinsic Dimensionality is on Arxiv.