Test-time scaling refers to scaling up test-time compute to improve performance of a language model. While there are many methods of doing so, one such method is large reasoning models.