Language Models can Self-Improve at State-Value Estimation for Better Search | Read Paper on Bytez