1 paper
Lijie Yang, Hongyin Luo, Jiawei Zhao +2
Test-time compute scaling is a primary driver of performance in large reasoning models (LRMs), but extreme inefficiency bounds current approaches, shifting the critical question fr…