pith. sign in

arxiv: 1706.05350 · v1 · pith:NNWZ43PUnew · submitted 2017-06-16 · 💻 cs.LG · stat.ML

L2 Regularization versus Batch and Weight Normalization

classification 💻 cs.LG stat.ML
keywords normalizationregularizationbatchinfluencelearningnetworksneuralrate
0
0 comments X
read the original abstract

Batch Normalization is a commonly used trick to improve the training of deep neural networks. These neural networks use L2 regularization, also called weight decay, ostensibly to prevent overfitting. However, we show that L2 regularization has no regularizing effect when combined with normalization. Instead, regularization has an influence on the scale of weights, and thereby on the effective learning rate. We investigate this dependence, both in theory, and experimentally. We show that popular optimization methods such as ADAM only partially eliminate the influence of normalization on the learning rate. This leads to a discussion on other ways to mitigate this issue.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 19 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Dead-Direction Conditioners: Gauge-Equivariant Preconditioning for Deep Networks

    cs.LG 2026-06 unverdicted novelty 7.0

    Dead-Direction Conditioners provide gauge-equivariant preconditioning by conditioning optimizer state on symmetry orbits, yielding improved resistance to over-training collapse and higher detection of dead directions ...

  2. Beyond Structural Symmetries: Linear Mode Connectivity via Neuron Identifiability

    cs.LG 2026-06 unverdicted novelty 7.0

    Neural networks admit large families of approximately equivalent solutions via neuron identifiability even without structural symmetry, enabling linear low-loss merging paths without prior alignment.

  3. Scalable Agentic Reasoning for Designing Biologics Targeting Intrinsically Disordered Proteins

    q-bio.QM 2025-12 unverdicted novelty 7.0

    StructBioReasoner is a scalable multi-agent system that designs IDP-targeting biologics, with over 50% of 787 candidates for Der f 21 showing better binding free energy than human-designed references.

  4. How does the optimizer implicitly bias the model merging loss landscape?

    cs.LG 2025-10 unverdicted novelty 7.0

    Effective noise scale non-monotonically governs model merging success with an optimum, unifying effects of learning rate, weight decay, batch size, and augmentation on the loss landscape.

  5. Progressive Growing of GANs for Improved Quality, Stability, and Variation

    cs.NE 2017-10 accept novelty 7.0

    Progressive growing stabilizes GAN training to produce high-resolution images of unprecedented quality and achieves a record unsupervised inception score of 8.80 on CIFAR10.

  6. On the Nonlinearity of Learning Rate Scaling for LLM Training

    cs.LG 2026-06 unverdicted novelty 6.0

    Optimal learning rate for models from 22M to 707M parameters shows nonlinear upward curvature with scale that disappears under effective learning rate and data-scale extrapolation.

  7. Improving Neural Network Training by Decoupling the Magnitude and Direction of Weight Vectors

    cs.LG 2026-06 unverdicted novelty 6.0

    MD Decoupling factorizes weights into fixed-norm directions and learnable per-row/column magnitudes updated at independent rates, improving Adam and Muon training stability and scale transfer without weight decay or warmup.

  8. Muown Implicitly Performs Angular Step-size Decay

    cs.LG 2026-06 conditional novelty 6.0

    Muown's update implicitly decays angular step size via magnitude modulation; AngularMuown decouples and schedules angular steps explicitly, yielding better empirical results.

  9. Weibull Weight-Scale Parameter Evolution under AdamW Training Dynamics

    cs.LG 2026-06 accept novelty 6.0

    Derives three-force decomposition of squared weight norm under AdamW and validates it on Pythia-70M models, plus spline recovery of alignment force from checkpoints.

  10. Preserving Plasticity in Continual Learning via Dynamical Isometry

    cs.LG 2026-06 unverdicted novelty 6.0

    Dynamical isometry (Jacobian singular values near 1) preserves plasticity in continual learning; an isometry-promoting regularizer and decoupled AdamO optimizer match or beat prior methods on supervised and RL benchmarks.

  11. Does Weight Decay Enhance Training Stability?

    cs.LG 2026-05 conditional novelty 6.0

    Weight decay slows progressive sharpening at the edge of stability, inducing damped oscillations in CNNs and a phase transition to sub-2/η sharpness in MLPs driven by parameter-sharpness gradient alignment, yielding m...

  12. Demystifying Manifold Constraints in LLM Pre-training

    cs.LG 2026-05 unverdicted novelty 6.0

    Manifold constraints via the new MACRO optimizer independently bound activation scales and enforce rotational equilibrium in LLM pre-training, subsuming RMS normalization and decoupled weight decay while delivering co...

  13. FlashSAC: Fast and Stable Off-Policy Reinforcement Learning for High-Dimensional Robot Control

    cs.LG 2026-04 unverdicted novelty 6.0

    FlashSAC improves training speed and final performance of off-policy RL on high-dimensional robot tasks by reducing update frequency, increasing model scale, and bounding norms to limit critic error accumulation.

  14. FlashSAC: Fast and Stable Off-Policy Reinforcement Learning for High-Dimensional Robot Control

    cs.LG 2026-04 unverdicted novelty 6.0

    FlashSAC scales up Soft Actor-Critic with fewer updates, larger models, higher data throughput, and norm bounds to deliver faster, more stable training than PPO on high-dimensional robot control tasks across dozens of...

  15. Can Stationary Distributions of Scale-Invariant Neural Networks Be Described by the Thermodynamics of an Ideal Gas?

    cs.LG 2025-11 unverdicted novelty 6.0

    A thermodynamic framework maps SGD stationary distributions in scale-invariant networks to ideal-gas behavior, with training hyperparameters acting as thermodynamic variables.

  16. Anytime Training with Schedule-Free Spectral Optimization

    cs.LG 2026-05 unverdicted novelty 5.0

    SF-NorMuon is a new schedule-free spectral optimizer that closes the gap with tuned AdamW on 125M-772M parameter models across 1-8x Chinchilla horizons while providing stationarity guarantees.

  17. ScheduleFree+: Scaling Learning-Rate-Free & Schedule-Free Learning to Large Language Models

    cs.LG 2026-05 unverdicted novelty 5.0

    ScheduleFree+ scales schedule-free learning to LLMs with fixes for large batches and models, outperforming Warmup-Stable-Decay schedules by up to 31% at 1000 tokens per parameter.

  18. XQCfD: Accelerating Fast Actor-Critic Algorithms with Prior Data and Prior Policies

    cs.LG 2026-05 unverdicted novelty 5.0

    XQCfD accelerates actor-critic RL by using prior data, pretrained policies, and stationary architectures to achieve state-of-the-art results on Adroit, Robomimic, and MimicGen manipulation benchmarks with low update-t...

  19. Adaptive Norm-Based Regularization for Neural Networks

    stat.ML 2026-04 unverdicted novelty 5.0

    Covariance-aware ridge and combined l1-l2 regularizers for neural networks yield better predictive performance and complexity control than standard penalties in simulations and applications to cooling-load prediction ...