Abstract: Current state-of-the-art plug-and-play countermeasures for mitigating adversarial examples (i.e., purification and detection) exhibit several fatal limitations, impeding their deployment in ...

This project contains implementations of simple neural network models, including training scripts for PyTorch and Lightning frameworks. The goal is to provide a modular, easy-to-understand codebase ...

The Model Context Protocol (MCP) is a standardized way for AI applications and agents to connect with data sources and tools. It allows AI models like Claude to interact with your local system, ...

Abstract: Force-sensitive manipulation is essential for tasks such as cleaning, polishing, and surgical assistance, yet it remains challenging due to complex contact dynamics and the need for ...