Network Sensitivity

Network sensitivity, encompassing the susceptibility of network models (including neural networks and language models) to variations in input, parameters, or training data, is a crucial area of research aiming to improve model robustness and reliability. Current investigations focus on quantifying sensitivity through various metrics, analyzing its relationship to model performance and identifying sources of oversensitivity in different architectures (e.g., CNNs, Transformers, LLMs). Understanding and mitigating network sensitivity is vital for enhancing the trustworthiness and generalizability of AI systems across diverse applications, from medical image analysis to natural language processing.