Today's large language models tend to have large parameters, and consumer-grade computers are slow to do simple inference, let alone train a model from scratch. The goal of this project is to organize ...

Abstract: Deep code models are vulnerable to adversarial attacks, making it possible for semantically identical inputs to trigger different responses. Current black-box attack methods typically ...

Abstract: A data-secure and cost-efficient Personalized Travel Recommendation (PTR) is necessary to develop urban intelligence in transportation. Although prevalent machine learning-based methods have ...