MS student, The University of Tokyo
1 paper at NeurIPS 2025
Revisiting layer-wise PTQ, propagating quantization errors enhances existing methods.