1 paper
Yujun Lin, Zhekai Zhang, Song Han
Modern tensor applications, especially foundation models and generative AI applications require multiple input modalities (both vision and language), which increases the demand for…