TY  - RPRT
TI  - LoRA-MGPO: Mitigating Double Descent in Low-Rank Adaptation via Momentum-Guided Perturbation Optimization
AU  - Yupeng Chang
AU  - Chenlu Guo
AU  - Yi Chang
AU  - Yuan Wu
PY  - 2025
UR  - https://arxiv.org/abs/2502.14538
ID  - 2502.14538
ER  -