AI SPR LU Luka Can Compute Buy Wisdom? In the current era where LLM Test-Time Scaling has become a dominant paradigm, we often fall into a myth: that simply increasing the number…