Trained models as described in the paper "ReSearch: Learning to Reason with Search for LLMs via Reinforcement Learning"