作者位次 #3/5ICML 2026Poster· Deep Learning / Large Language Models
Preserve-Then-Quantize: Balancing Rank Budgets for Quantization Error Reconstruction in LLMs
Yoonjun Cho · Dongjae Jeon · Soeun Kim · Moongyu Jeon · Albert No
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。
Yoonjun Cho · Dongjae Jeon · Soeun Kim · Moongyu Jeon · Albert No
Wonje Jeung · Sangyeon Yoon · Hyesoo Hong · Soeun Kim · Seungju Han · Youngjae Yu · Albert No
Machine unlearning aims to remove “forget” data while preserving knowledge from the “retain” data, yet a fundamental question arises when the two share content. By definition, an unlearned model should be indistinguishable from a model retrained solely on the retain set, which implies that shared...