TY  - RPRT
TI  - Dynamically Weighted Momentum with Adaptive Step Sizes for Efficient Deep Network Training
AU  - Zhifeng Wang
AU  - Longlong Li
AU  - Chunyan Zeng
PY  - 2025
UR  - https://arxiv.org/abs/2510.25042
ID  - 2510.25042
ER  -