Dropout Universality: Scaling Laws and Optimal Scheduling at the Edge-of-Chaos

• URL: http://arxiv.org/abs/2605.21648v1
• Date: Wed, 20 May 2026 19:00:02 GMT
• Title: Dropout Universality: Scaling Laws and Optimal Scheduling at the Edge-of-Chaos
• Abstract summary: We develop a mean-field theory of dropout as a perturbation of critical signal propagation at the edge of chaos.<n>We derive critical and crossover scaling laws for correlation decay and establish that smooth activations and kinked, ReLU-like activations constitute distinct classes.
• Score: 0.2864713389096699
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We develop a mean-field theory of dropout as a perturbation of critical signal propagation at the edge of chaos. Dropout shifts the perfect-alignment fixed point, making the depth scale for information propagation finite even at critical initialization. We derive critical and crossover scaling laws for correlation decay and establish that smooth activations and kinked, ReLU-like activations constitute distinct universality classes, with different critical exponents and a universal two-parameter scaling collapse in detuning and dropout strength. The distinction traces to the analytic structure of the correlation map: smooth activations admit a Taylor expansion near perfect alignment, while kinked activations develop a branch point with universal non-analyticity. As a corollary, the framework yields saturated dropout profiles under fixed budget; a rank-flow tie-breaker then selects front-loaded schedules, substantially reducing held-out test loss at no extra computational cost, with accuracy gains as a consistent secondary effect. We test the predictions in MLPs and Vision Transformers and discuss CNN/ResNet extensions.

Related papers

• Across the Loss Landscape with Progressive Growth [51.366966420881646]
  We show how grow-and-optimize strategies to train toward flatter regions by viewing growth as progressive relaxation.<n>We empirically validate these predictions in controlled toy basins and in a realistic ResNet/CI-100 setting.<n>Although progressive subspace growth reliably produces flatter solutions, curvature reductions do not universally translate into improved test performance, highlighting subtleties in the flatness-generalization connection.
  arXiv  Detail & Related papers  (2026-08-25T13:50:48Z)
• NICE: Scale-Stable Perturbations for Graph Neural Network Explanations via Noise Corruption [67.57304101271096]
  Element-wise Masking (EM) suppresses edge-induced messages toward zero.<n>Noise Corruption (NC) perturbs each message through matched-norm random-direction corruption.<n>NICE, a Noise Corruption-based explanation framework, learns a Restoration Boundary (SRB) under NC-induced uncertainty.
  arXiv  Detail & Related papers  (2026-08-17T03:08:14Z)
• Drift and Dependence: Layer-wise Information-Theoretic Bounds for Replay-Based Continual Learning [41.371439825328544]
  Continual learning must absorb new tasks without erasing old ones, and replay---mixing a small buffer of past examples into current training---is among the most effective remedies for catastrophic forgetting.<n>We develop a layer-wise information-theoretic framework that separates these effects at every depth.<n>Controlled and benchmark experiments confirm the predicted memory scaling, the interior funnel, and the alignment signal's link to forgetting.
  arXiv  Detail & Related papers  (2026-08-12T06:01:49Z)
• Distributed Optimization with Streaming Data: A Temporal Weighting Perspective [26.702871313273942]
  We study decentralized optimization from streaming data through a structured time-varying formulation.<n>We analyze multi-iteration decentralized first-order methods, including decentralized gradient descent.<n>We specialize our analysis to uniform and exponentially discounted weights, as well as their finite-memory emphwindowed counterparts.
  arXiv  Detail & Related papers  (2026-08-10T13:02:58Z)
• When Do Autoregressive Sequence Models Forecast Physical Wavefields? A Controlled Study on Synthetic Seismograms [43.8784307709823]
  Long-horizon autoregressive forecasting of physical signals is limited by error accumulation.<n>We ask when such rollout stays stable using synthetic three-component seismograms as a physically structured testbed.
  arXiv  Detail & Related papers  (2026-06-09T13:46:14Z)
• The Structural Origin of Attention Sink: Variance Discrepancy, Super Neurons, and Dimension Disparity [22.040532283910522]
  This work provides a textitmechanistic explanation for the phenomenon.<n>We trace its root to the value aggregation process inherent in self-attention.<n>As a proof of concept, we propose textithead-wise RMSNorm, an architectural modification that stabilizes value aggregation outputs during pre-training.
  arXiv  Detail & Related papers  (2026-05-07T17:28:55Z)
• The Geometric Inductive Bias of Grokking: Bypassing Phase Transitions via Architectural Topology [0.0]
  We study grokking - delayed generalization in Transformers trained on cyclic modular addition (Zp)<n>We identify two independent structural factors in standard Transformers: representational magnitude and data-dependent attention routing.
  arXiv  Detail & Related papers  (2026-03-05T14:41:01Z)
• Generalizing GNNs with Tokenized Mixture of Experts [75.8310720413187]
  We show that improving stability requires reducing reliance on shift-sensitive features, leaving an irreducible worst-case generalization floor.<n>We propose STEM-GNN, a pretrain-then-finetune framework with a mixture-of-experts encoder for diverse computation paths.<n>Across nine node, link, and graph benchmarks, STEM-GNN achieves a stronger three-way balance, improving robustness to degree/homophily shifts and to feature/edge corruptions while remaining competitive on clean graphs.
  arXiv  Detail & Related papers  (2026-02-09T22:48:30Z)
• MirrorLA: Reflecting Feature Map for Vision Linear Attention [49.41670925034762]
  Linear attention significantly reduces the computational complexity of Transformers from quadratic to linear, yet it consistently lags behind softmax-based attention in performance.<n>We propose MirrorLA, a geometric framework that substitutes passive truncation with active reorientation.<n>MirrorLA achieves state-of-the-art performance across standard benchmarks, demonstrating that strictly linear efficiency can be achieved without compromising representational fidelity.
  arXiv  Detail & Related papers  (2026-02-04T09:14:09Z)
• A Unified View of Attention and Residual Sinks: Outlier-Driven Rescaling is Essential for Transformer Training [86.64715217940274]
  Outliers function jointly with normalization.<n>Outliers serve more as rescale factors rather than contributors.<n>Outliers can be absorbed into learnable parameters or mitigated via explicit gated rescaling.
  arXiv  Detail & Related papers  (2026-01-30T13:29:45Z)
• The Procrustean Bed of Time Series: The Optimization Bias of Point-wise Loss [53.542743390809356]
  This paper aims to provide a first-principles analysis of the Expectation of Optimization Bias (EOB)<n>Our analysis reveals a fundamental paradigm paradox: the more deterministic and structured the time series, the more severe the bias by point-wise loss function.<n>We present a concrete solution that simultaneously achieves both principles via DFT or DWT.
  arXiv  Detail & Related papers  (2025-12-21T06:08:22Z)
• Localization from Infinitesimal Kinetic Grading: Critical Scaling and Kibble-Zurek Universality [0.0]
  We study a one-dimensional lattice model with site-dependent nearest-neighbor hopping amplitudes that follow a power-law profile.<n>In the thermodynamic limit, the ground state becomes localized as $|| to 0$, signaling the presence of a critical point characterized by a diverging localization length.<n>Our results demonstrate a clean, disorder-free route to localization and provide a tunable platform relevant to photonic lattices and ultracold atom arrays with engineered hopping profiles.
  arXiv  Detail & Related papers  (2025-12-16T17:26:06Z)
• Revisiting Zeroth-Order Optimization: Minimum-Variance Two-Point Estimators and Directionally Aligned Perturbations [57.179679246370114]
  We identify the distribution of random perturbations that minimizes the estimator's variance as the perturbation stepsize tends to zero.<n>Our findings reveal that such desired perturbations can align directionally with the true gradient, instead of maintaining a fixed length.
  arXiv  Detail & Related papers  (2025-10-22T19:06:39Z)
• Implicit Bias and Fast Convergence Rates for Self-attention [26.766649949420746]
  We study the fundamental optimization principles of self-attention, the defining mechanism of transformers.<n>We analyze the implicit bias of gradient-baseds in a self-attention layer with a decoder in a linear classification.
  arXiv  Detail & Related papers  (2024-02-08T15:15:09Z)
• Stable Nonconvex-Nonconcave Training via Linear Interpolation [51.668052890249726]
  This paper presents a theoretical analysis of linearahead as a principled method for stabilizing (large-scale) neural network training. We argue that instabilities in the optimization process are often caused by the nonmonotonicity of the loss landscape and show how linear can help by leveraging the theory of nonexpansive operators.
  arXiv  Detail & Related papers  (2023-10-20T12:45:12Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.