Spread-Out Regularization in Matryoshka Text Embeddings: A Retrieval-Focused Analysis
This paper presents a controlled analysis of whether spread-out regularization, a batch-level loss that penalizes high pairwise similarity within a batch, can improve representation-space utilization in Matryoshka Representation Learning settings and shows that spread-out loss is a retrieval-biased regularizer.
Zeqi Zhou, Stephen Bach
· 0 citations