DyRA: Dynamic Residual Approximation for Efficient Matrix Multiplication in DNNs

תקציר מקורי באנגליתarXiv:2610.02882v1 Announce Type: new Abstract: Large-scale foundation models achieve strong performance across diverse tasks, but their size makes inference costly, largely due to dense matrix multiplications. Prior work reduces this cost by replacing dense weight matrices with efficient structured forms such as low-rank factorizations. However, these methods approximate weights rather than the output activations that determine inference accuracy. Consequently, small weight-space errors can be amplified by input activations, producing large output errors. In this work, we propose DyRA, an input-adaptive method that improves structured matrix multiplication approximation by correcting residual output errors during inference. We show that matrix multiplication can be approximated more effec