AI RESEARCH
GD-FPS: Growth-Driven Feedforward Parameter Selection for Efficient Fine-Tuning
arXiv CS.LG
•
ArXi:2510.27359v2 Announce Type: replace-cross Parameter-Efficient Fine-Tuning (PEFT) has emerged as a key strategy for adapting large-scale pre-trained models to downstream tasks, but existing approaches face notable limitations. Addition-based methods, such as Adapters,