G^2PTQ: Improving LLM Post-Training Quantization with Generalized Gradient Compensation ▲ 2 💬 1 2026-09-30 arXiv HF 原文 GitHub ★2 摘要 (无摘要)