Learning to Perturb Hidden Representations for Generalizable Deep Learning

• URL: http://arxiv.org/abs/2605.29525v1
• Date: Thu, 28 May 2026 07:41:39 GMT
• Title: Learning to Perturb Hidden Representations for Generalizable Deep Learning
• Abstract summary: We show that Dropout, Manifold Mixup, adversarial feature perturbation, and related methods all impose specific forms of activation perturbation but with class-agnostic or random strategies.<n>We propose Learning to Perturb Activations (LPA), which adaptively perturbs activations at a selected hidden layer with class-level perturbations learned via PGD.
• Score: 3.2608238154676843
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Deep neural networks process data through a cascade of representations: input features, hidden activations, logits, and loss. While perturbations at the input, logit, and label levels have been systematically studied, the intermediate hidden activations, which constitute the bulk of the network's computation, have received no unified perturbation analysis. In this paper, we establish a unified framework for hidden activation perturbation, revealing that Dropout, Manifold Mixup, adversarial feature perturbation, and related methods all impose specific forms of activation perturbation but with class-agnostic or random strategies. We conjecture that expansive perturbation (increasing activation norm) acts as positive augmentation, while contractive perturbation (decreasing activation norm) acts as negative augmentation, and that the perturbation layer determines whether the effect resembles input-level augmentation (shallow layers) or logit-level manipulation (deep layers). We propose Learning to Perturb Activations (LPA), which adaptively perturbs activations at a selected hidden layer with class-level perturbations learned via PGD. We further provide theoretical analysis connecting activation perturbation to flat minima and perturbation amplification through layers. Experiments on balanced classification, long-tail classification, and domain generalization demonstrate that LPA consistently outperforms existing methods and provides complementary benefits to logit perturbation methods such as LPL.

Related papers

• Sparse Inter-Layer Dependencies of Transformer FFN Neurons [32.00544060412007]
  We study whether the activation of an FFN neuron can be explained by a sparse set of preceding neuron activations and attention outputs.<n>We introduce a training-free attribution method that estimates the relative influence of upstream neurons and attention outputs on a target neuron's activation.
  arXiv  Detail & Related papers  (2026-07-13T15:04:50Z)
• Beyond Layer Importance in Layer-wise Sparsity: An Inter-Layer Perturbation-Absorption Perspective [22.968807193719044]
  Layer-wise redundancy in large language models has established non-uniform sparsity allocation across layers as the standard pruning approach for efficient compression.<n>We characterize this property through controlled perturbation experiments.<n>We propose absorption-aware correction, an augmentation that improves outlier and AlphaPruning by reducing perplexity by 7.13% and boosting zero-shot accuracy by 1.02% across multiple model families at 70% sparsity.
  arXiv  Detail & Related papers  (2026-06-13T07:16:16Z)
• Gradient Perturbation: Learning to Perturb Gradients for Adaptive Training [3.2608238154676843]
  gradient perturbation has received little systematic investigation.<n>We propose Learning to Perturb Gradients (LPG), which adaptively perturbs logit-level gradients at the class level to achieve category-aware training.<n> Experiments on balanced classification, long-tail classification, and noisy label learning demonstrate LPG consistently outperforms existing methods and can be combined with them as a plug-in module.
  arXiv  Detail & Related papers  (2026-05-28T07:19:36Z)
• A Single Layer to Explain Them All:Understanding Massive Activations in Large Language Models [56.03096341469333]
  We investigate the origins of massive activations in large language models (LLMs)<n>We identify a specific layer named the textbfMassive Emergence Layer (ME Layer), that is consistently observed across model families.<n>We show that, within the ME Layer both the RMSNorm and the FFN parameters jointly contribute to the emergence of massive activations.<n>Motivated by this limitation, we propose a simple and effective method to reduce the rigidity of the massive activation token.
  arXiv  Detail & Related papers  (2026-05-08T21:37:27Z)
• When Does Sparsity Mitigate the Curse of Depth in LLMs [53.137717161619484]
  We show that sparsity acts as a regulator of variance propagation and thereby improves depth utilization.<n>Our results reveal sparsity as a key yet previously overlooked mechanism for effective depth scaling in large language models.
  arXiv  Detail & Related papers  (2026-03-16T15:04:16Z)
• From Latent Signals to Reflection Behavior: Tracing Meta-Cognitive Activation Trajectory in R1-Style LLMs [48.33546389897804]
  R1-style LLMs have attracted growing attention for their capacity for self-reflection, yet the internal mechanisms underlying such behavior remain unclear.<n>Using the logit lens to read out token-level semantics, we uncover a structured progression.<n>Our findings suggest a human-like metacognitive process-progressing from latent monitoring, to discourse-level regulation, and to finally overt self-reflection.
  arXiv  Detail & Related papers  (2026-02-02T11:58:24Z)
• Probe-Free Low-Rank Activation Intervention [26.502232859901167]
  Inference-time interventions that edit the hidden activations have shown promising results in steering the LMs towards desirable generations.<n>This paper proposes a probe-free intervention method FLORAIN for all attention heads in a specific activation layer.
  arXiv  Detail & Related papers  (2025-02-06T13:03:05Z)
• Learning Neural Networks with Sparse Activations [42.88109060676769]
  In transformer networks, the activations in the hidden layer of this block tend to be extremely sparse on any given input. Unlike traditional forms of sparsity, where there are neurons/weights which can be deleted from the network, this form of em activation sparsity appears to be harder to exploit. We present a variety of results showing that classes of functions do lead to provable computational and statistical advantages over their non-sparse counterparts.
  arXiv  Detail & Related papers  (2024-06-26T00:11:13Z)
• Calibrating Undisciplined Over-Smoothing in Transformer for Weakly Supervised Semantic Segmentation [51.14107156747967]
  Weakly supervised semantic segmentation (WSSS) has attracted considerable attention because it requires fewer annotations than fully supervised approaches.<n>We propose an Adaptive Re-Activation Mechanism (AReAM) to control deep-level attention to undisciplined over-smoothing.<n>AReAM substantially improves segmentation performance compared with existing WSSS methods, reducing noise while sharpening focus on relevant semantic regions.
  arXiv  Detail & Related papers  (2023-05-04T19:11:33Z)
• MCDAL: Maximum Classifier Discrepancy for Active Learning [74.73133545019877]
  Recent state-of-the-art active learning methods have mostly leveraged Generative Adversarial Networks (GAN) for sample acquisition. We propose in this paper a novel active learning framework that we call Maximum Discrepancy for Active Learning (MCDAL) In particular, we utilize two auxiliary classification layers that learn tighter decision boundaries by maximizing the discrepancies among them.
  arXiv  Detail & Related papers  (2021-07-23T06:57:08Z)
• The Impact of Activation Sparsity on Overfitting in Convolutional Neural Networks [1.9424280683610138]
  Overfitting is one of the fundamental challenges when training convolutional neural networks. In this study we introduce a perplexity-based sparsity definition to derive and visualise layer-wise activation measures.
  arXiv  Detail & Related papers  (2021-04-13T12:55:37Z)
• Interpreting Deep Neural Networks with Relative Sectional Propagation by Analyzing Comparative Gradients and Hostile Activations [37.11665902583138]
  We propose a new attribution method, Relative Sectional Propagation (RSP), for decomposing the output predictions of Deep Neural Networks (DNNs) We define hostile factor as an element that interferes with finding the attributions of the target and propagates it in a distinguishable way to overcome the non-suppressed nature of activated neurons. Our method makes it possible to decompose the predictions of DNNs with clearer class-discriminativeness and detailed elucidations of activation neurons compared to the conventional attribution methods.
  arXiv  Detail & Related papers  (2020-12-07T03:11:07Z)
• Extreme Memorization via Scale of Initialization [72.78162454173803]
  We construct an experimental setup in which changing the scale of initialization strongly impacts the implicit regularization induced by SGD. We find that the extent and manner in which generalization ability is affected depends on the activation and loss function used. In the case of the homogeneous ReLU activation, we show that this behavior can be attributed to the loss function.
  arXiv  Detail & Related papers  (2020-08-31T04:53:11Z)

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.