AI RESEARCH

EoRA: Fine-tuning-free Compensation for Compressed LLM with Eigenspace Low-Rank Approximation

arXiv CS.AI

ArXi:2410.21271v5 Announce Type: replace-cross While post-