Skip to main content
    Skip to main contentSkip to navigationSkip to footer
    Artificial Intelligence

    Spectral Normalization

    Also known as:
    Spectral Norm
    SN
    Updated: 2/12/2026

    Spectral Normalization constrains the Lipschitz constant of network layers by normalizing with the largest singular value – standard stabilization in GANs.

    Quick Summary

    Spectral Normalization constrains the Lipschitz constant via the largest singular value – the standard stabilization for GANs.

    Explanation

    W_SN = W / σ(W), where σ(W) is the largest singular value (efficiently computed via power iteration). Limits the "aggressiveness" of the discriminator in GANs and stabilizes training without gradient penalty.

    Marketing Relevance

    Standard stabilization technique in GANs (SNGAN, BigGAN, StyleGAN) and increasingly in diffusion models.

    Origin & History

    Miyato et al. (2018) introduced Spectral Normalization for GANs (SNGAN). BigGAN (2018) and StyleGAN (2019) adopted the technique. Today also common in diffusion models (e.g., Stable Diffusion U-Net).

    Comparisons & Differences

    Spectral Normalization vs. Gradient Penalty (WGAN-GP)

    Gradient Penalty computes gradient norm per batch (expensive); Spectral Norm normalizes weights directly (faster, simpler).

    Related Services

    Related Terms

    Weight NormalizationBatch Normalizationgenerative-adversarial-network
    👋Questions? Chat with us!