Not All Bits Are Equal: Scale-Dependent Memory Optimization Strategies for Reasoning Models
机构 * KRAFTON ; University of Wisconsin–Madison(威斯康星大学麦迪逊分校) ; UC Berkeley(加州大学伯克利分校) ; Microsoft Research(微软研究院)
Comments 20 pages, 12 figures
高校专区
机构 * KRAFTON ; University of Wisconsin–Madison(威斯康星大学麦迪逊分校) ; UC Berkeley(加州大学伯克利分校) ; Microsoft Research(微软研究院)
Comments 20 pages, 12 figures
机构 * University of Wisconsin-Madison(威斯康星大学麦迪逊分校) ; Virginia Tech(弗吉尼亚理工学院)
Journal ref 39th Conference on Neural Information Processing Systems (NeurIPS 2025) Workshop: New Perspectives in Advancing Graph Machine Learning