When Lower Reconstruction Loss Hurts: Distributionally Robust Refinement for Low-Bit LLM Quantization
arXiv:2610.11226v1 Announce Type: new Abstract: Weight-only post-training quantization (PTQ) relies heavily on reconstruction loss minimization to preserve model quality at low precision. We show that the weights favored by minimizing this...
Read more: https://arxiv.org/abs/2610.11226
Scan with your phone
Scan with your phone