arXiv cs.LG
· Papers
Beyond Rotations: AuroOFT for Expressive Quantized Orthogonal Fine-Tuning
arXiv:2608.05253v1 Announce Type: new Abstract: Quantized orthogonal fine-tuning (qoft) enables parameter-efficient adaptation of low-bit language models by learning structured activation rotations before frozen quantized weights. However, its task-specific updates remain constrained to linear orthogonal transformation