English

Probabilistic Contrastive Loss for Self-Supervised Learning

Machine Learning 2021-12-06 v1

Abstract

This paper proposes a probabilistic contrastive loss function for self-supervised learning. The well-known contrastive loss is deterministic and involves a temperature hyperparameter that scales the inner product between two normed feature embeddings. By reinterpreting the temperature hyperparameter as a quantity related to the radius of the hypersphere, we derive a new loss function that involves a confidence measure which quantifies uncertainty in a mathematically grounding manner. Some intriguing properties of the proposed loss function are empirically demonstrated, which agree with human-like predictions. We believe the present work brings up a new prospective to the area of contrastive learning.

Keywords

Cite

@article{arxiv.2112.01642,
  title  = {Probabilistic Contrastive Loss for Self-Supervised Learning},
  author = {Shen Li and Jianqing Xu and Bryan Hooi},
  journal= {arXiv preprint arXiv:2112.01642},
  year   = {2021}
}
R2 v1 2026-06-24T08:02:32.304Z