Showing cs.LGShow all
2 papers · 1 filter
cs.LG2025
Decipher the Modality Gap in Multimodal Contrastive Learning: From Convergent Representations to Pairwise Alignment
Lingjie Yi, Raphael Douady, Chao Chen
Multimodal contrastive learning (MCL) aims to embed data from different modalities in a shared embedding space. However, empirical evidence shows that representations from differen…
cs.LG2024
A Theoretical Study of Neural Network Expressive Power via Manifold Topology
Jiachen Yao, Mayank Goswami, Chao Chen
A prevalent assumption regarding real-world data is that it lies on or close to a low-dimensional manifold. When deploying a neural network on data manifolds, the required size, i.…