An Information-Theoretic Perspective on Variance-Invariance-Covariance Regularization

2023/03/01 by Ravid Shwartz-Ziv, Shwartz-Ziv, Ravid, Randall Balestriero +7 · 2 citations
Computer Science · #Artificial Intelligence (cs.AI) #Domain Adaptation and Few-Shot Learning #FOS: Computer and information sciences #Information Theory (cs.IT) #Machine Learning and ELM #Neural Networks and Applications

paper · pdf · doi:10.48550/arxiv.2303.00633

openalex publication_date 2023/03/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Variance-Invariance-Covariance Regularization (VICReg) is a self-supervised learning (SSL) method that has shown promising results on a variety of tasks. However, the fundamental mechanisms underlying VICReg remain unexplored. In this paper, we present an information-theoretic perspective on the VICReg objective. We begin by deriving information-theoretic quantities for deterministic networks as an alternative to unrealistic stochastic network assumptions. We then relate the optimization of the VICReg objective to mutual information optimization, highlighting underlying assumptions and facilitating a constructive comparison with other SSL algorithms and derive a generalization bound for VICReg, revealing its inherent advantages for downstream tasks. Building on these results, we introduce a family of SSL methods derived from information-theoretic principles that outperform existing SSL techniques.

Cited by

• Self-Supervised Learning with Gaussian Processes
• The Alignment Bottleneck

Related

• What Do We Maximize in Self-Supervised Learning?
• VICReg: Variance-Invariance-Covariance Regularization for\n Self-Supervised Learning
• Variance Covariance Regularization Enforces Pairwise Independence in Self-Supervised Representations
• Explicit Mutual Information Maximization for Self-Supervised Learning
• Self-Supervised Representation Learning as Mutual Information Maximization