Training-Memory, Storage, and Accuracy Trade-Offs of Magnitude-Selected Subspace and Multi-Locus Deletion Tuning
Deletion tuning adapts frozen language models via binary masks $\mathbf{W}' = \mathbf{W} \odot \mathbf{M}$ ($\mathbf{M} \in \{0,1\}^{d_{\text{out}} \times d_{\t