Reasoning

Predicting LLM Reasoning Performance with Small Proxy Model

AdaSTaR: Adaptive Data Sampling for Training Self-Taught Reasoners