作者位次 #3/5NeurIPS 2026Poster
SVDP: Training-Free Contextual Sparsity Predictors for Fast LLM Inference
Georgii Serbin · Kirill Koshkin · Zhongao Sun · Anastasiya Bistrigova · Constantine Korikov
效率与系统OpenReview
首次获取该候选人的完整档案及 StartLux 评估共 20 积分；每位用户单独解锁，之后查看免费。主动刷新 20 积分。
实时查找电子邮箱，费用不超过20积分；未找到邮箱不扣费。
Georgii Serbin · Kirill Koshkin · Zhongao Sun · Anastasiya Bistrigova · Constantine Korikov
Manyi Zhang · Ji-Fu Li · Zhongao Sun · Haoli Bai · Hui-Ling Zhen · Zhenhua Dong · Xianzhi Yu
Microscaling Floating-Point (MXFP) has emerged as a promising low-precision format for large language models (LLMs). Despite various post-training quantization (PTQ) algorithms being proposed, they mostly focus on integer quantization, while their applicability and behavior under MXFP formats rem...