Our paper Re-FORC: Adaptive Reward Prediction for Efficient Chain-of-Thought Reasoning is now on arXiv!