This paper proposes an approximation algorithm for streaming stochastic submodular maximization problem under a novel on-demand user requests senario

We propose an explainable and extendable framework to enhance deepfake detection via multimodal large-language models.

We introduce the Fair Minimum Labeling problem for designing temporally efficient and fair activation plans, prove tight hardness bounds, and present approximation algorithms with strong empirical results on fair multi-source learning.

We proposed a new data selection method for pretraining multilingual Large Language Models