Researchers are exploring various approaches to improve intrusion detection for internet of things (IoT) applications, including hybrid models that combine different deep learning techniques. These approaches aim to enhance the accuracy and interpretability of intrusion detection systems, with some incorporating explainable artificial intelligence and dimensionality reduction methods. The specific methods being investigated include combinations of BiLSTM, CNN, and other techniques, with applications extending to related areas such as smart maritime threat detection and DDoS classification.