Efficiency

Predicting LLM Reasoning Performance with Small Proxy Model