Pruned Model

Pruned models aim to reduce the size and computational cost of deep neural networks (DNNs) by removing less important parameters or neurons, while preserving accuracy. Current research focuses on developing efficient pruning algorithms, including structured pruning methods for large language models (LLMs) and vision transformers, and exploring techniques like optimal transport for mitigating security vulnerabilities in pruned models. This work is significant because it addresses the growing need for deploying DNNs on resource-constrained devices and improves the efficiency and security of various machine learning applications.