Reproduce “CORA: Per-Slice Coherent Orthogonal Rotation for SVD-based Low-Rank Adaptation” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Pengcheng Wang, Ziran Liu, Wei Wang, Wei Jiang
CORA adapts pretrained weights with parameter-efficient updates that rotate both singular-direction bases of row slices together and learn a shared layer scale. It uses top-r SVD components, with a minimum-perturbation characterization under assumptions. On LLaMA-2-7B commonsense reasoning, CORA rank 16 reports 82.16% accuracy with 6.6M trainable parameters, versus DoRA’s 79.7% with 57M; it remains below full fine-tuning at 83.5%.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.