Detecting deceptive behavior in LLMs is typically done post-hoc on outputs or by probing static activations.We instead treat deception as a dynamic process, a trajectory through the model's hidden-state space during inference.We capture layerwise activations at sparse "decision points" where the model is uncertain between competing tokens, forming activation trajectories for matched truthful vs. deceptive responses across strategic deception, sycophancy, instructed deception, and confabulation.Across GPT-2 and Llama variants, deceptive generation is associated with changes in trajectory geometry, but increases in path length are model and deception-type-dependent.Sycophancy shows the clearest signal, whereas instructed deception yields near-null signatures.With just 7 geometric features, a lightweight classifier achieves performance comparable to PCA-reduced probing at matched dimensionality for binary sycophancy detection and shows preliminary utility for 4-way deceptiontype classification.These findings indicate that trajectory-based monitoring can provide process-level signals associated with deceptive generation during inference, complementing methods that focus on endpoint activation states.

Intelligent sensing solutions bridge the gap between the physical world and the cyber-physical systems by digitizing the sensor data collected from sensor devices. Sensor cloud networks provide physical and virtual sensing device resources and enable uninterrupted intelligent solutions to end-users. Thanks to advancements in machine learning algorithms and big data, the automation of mundane tasks with artificial intelligence is becoming a reliable smart option. However, existing approaches based on centralized Machine Learning (ML) on sensor cloud networks fail to ensure data privacy. Moreover, centralized ML works with the pre-requisite to transfer the entire training dataset from end devices to a central server. To address this, we propose a Quantized Federated Learning (FL) based approach, called CloudFL, to ensure data privacy on end devices in a sensor cloud network. Our framework enables a personalized version of FL implementation and enhances privacy and security with cryptosystem tools to obfuscate the information of the FL process from unauthorized access. Furthermore, microservices of our approach provide software as a service implementation of FL with instances of cloud servers that require zero-touch on local data for training.

The Internet of Things (IoT) is made up of billions of physical devices connected to the Internet via networks that perform tasks independently with less human intervention. Such brilliant automation of mundane tasks requires a considerable amount of user data in digital format, which, in turn, makes IoT networks an open source of personally identifiable information data for malicious attackers to steal, manipulate, and perform nefarious activities. A huge interest has been developed over the past years in applying machine learning (ML)-assisted approaches in the IoT security space. However, the assumption in many current works is that big training data are widely available and transferable to the main server because data are born at the edge and are generated continuously by IoT devices. This is to say that classic ML works on the legacy set of entire data located on a central server, which makes it the least preferred option for domains with privacy concerns on user data. To address this issue, we propose the federated-learning (FL)-based anomaly detection approach to proactively recognize intrusion in IoT networks using decentralized on-device data. Our approach uses federated training rounds on gated recurrent units (GRUs) models and keeps the data intact on local IoT devices by sharing only the learned weights with the central server of FL. Also, the approach’s ensembler part aggregates the updates from multiple sources to optimize the global ML model’s accuracy. Our experimental results demonstrate that our approach outperforms the classic/centralized machine learning (non-FL) versions in securing the privacy of user data and provides an optimal accuracy rate in attack detection.

Billions of IoT devices are connected to networks all around us, enabling cyber-physical systems. These devices can carry and generate user-sensitive data, examples of such devices are smartwatches, medical equipment, and smart home gadgets. Individual IoT devices have some form of intrusion detection system integrated, but once they are all connected, a network threat to one device could mean a threat to many. IoT devices must have a robust intrusion detection system that would keep devices secure over a network. To aid with this, we provide a machine learning solution that adheres to Global Data Protection Regulation by keeping the user data secure locally on the IoT device itself. We propose a Federated Learning (FL) approach that capitalizes on a decentralized and collaborative way of training machine learning models. In this study, we practice federated learning technique to train and create a robust intrusion detection model for the security of IoT devices. We evaluate our proposed approach using three different use-cases to show the security enhancements that improve using the FL technique, resulting in a more reliable performance in this domain.

Hadoop Distributed File System (HDFS) is one of the widely used distributed file systems in big data analysis for frameworks such as Hadoop. HDFS allows one to manage large volumes of data using low-cost commodity hardware. However, vulnerabilities in HDFS can be exploited for nefarious activities. This reinforces the importance of ensuring robust security to facilitate file sharing in Hadoop as well as having a trusted mechanism to check the authenticity of shared files. This is the focus of this paper, where we aim to improve the security of HDFS using a blockchain-enabled approach (hereafter referred to as BlockHDFS). Specifically, the proposed BlockHDFS uses the enterprise-level Hyperledger Fabric platform to capitalize on files' metadata for building trusted data security and traceability in HDFS.