CrossQ: Improving Sample Efficiency in Deep Reinforcement Learning with Batch Normalization

CrossQ introduces a lightweight algorithm for continuous control tasks that enhances sample efficiency by leveraging Batch Normalization and eliminating target networks.