Randomly Initialized Networks Can Learn from Peer-to-Peer Consensus

• URL: http://arxiv.org/abs/2604.18390v1
• Date: Mon, 20 Apr 2026 15:13:39 GMT
• Title: Randomly Initialized Networks Can Learn from Peer-to-Peer Consensus
• Abstract summary: In self-supervised learning, self-distilled methods have shown impressive performance and learning representations useful for downstream tasks.<n>We explore the role of self-distillation within learning dynamics by training a group of randomly networks.<n>Our findings show that even this minimal setup can lead to learned representations with non-trivial improvements over a random baseline on downstream tasks.
• Score: 0.0
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: In self-supervised learning, self-distilled methods have shown impressive performance, learning representations useful for downstream tasks and even displaying emergent properties. However, state-of-the-art methods usually rely on ensembles of complex mechanisms, with many design choices that are empirically motivated and not well understood. In this work, we explore the role of self-distillation within learning dynamics. Specifically, we isolate the effect of self-distillation by training a group of randomly initialized networks, removing all other common components such as projectors, predictors, and even pretext tasks. Our findings show that even this minimal setup can lead to learned representations with non-trivial improvements over a random baseline on downstream tasks. We also demonstrate how this effect varies with different hyperparameters and present a short analysis of what is being learned by the models under this setup.

Related papers

• How to Probe: Simple Yet Effective Techniques for Improving Post-hoc Explanations [69.72654127617058]
  Post-hoc importance attribution methods are a popular tool for "explaining" Deep Neural Networks (DNNs)<n>In this work we bring forward empirical evidence that challenges this very notion.<n>We discover a strong dependency on and demonstrate that the training details of a pre-trained model's classification layer play a crucial role.
  arXiv  Detail & Related papers  (2025-03-01T22:25:11Z)
• Shortcut Learning Susceptibility in Vision Classifiers [11.599035626374409]
  Shortcut learning is where machine learning models exploit spurious correlations in data instead of capturing meaningful features.<n>This study introduces deliberate shortcuts into the dataset that are correlated with class labels both positionally and via intensity.<n>We evaluate susceptibility to shortcut learning across different learning rates.
  arXiv  Detail & Related papers  (2025-02-13T10:25:52Z)
• Deep Learning Through A Telescoping Lens: A Simple Model Provides Empirical Insights On Grokking, Gradient Boosting & Beyond [61.18736646013446]
  In pursuit of a deeper understanding of its surprising behaviors, we investigate the utility of a simple yet accurate model of a trained neural network. Across three case studies, we illustrate how it can be applied to derive new empirical insights on a diverse range of prominent phenomena.
  arXiv  Detail & Related papers  (2024-10-31T22:54:34Z)
• A Probabilistic Model Behind Self-Supervised Learning [53.64989127914936]
  In self-supervised learning (SSL), representations are learned via an auxiliary task without annotated labels. We present a generative latent variable model for self-supervised learning. We show that several families of discriminative SSL, including contrastive methods, induce a comparable distribution over representations.
  arXiv  Detail & Related papers  (2024-02-02T13:31:17Z)
• A Quantitative Approach to Predicting Representational Learning and Performance in Neural Networks [5.544128024203989]
  Key property of neural networks is how they learn to represent and manipulate input information in order to solve a task. We introduce a new pseudo-kernel based tool for analyzing and predicting learned representations.
  arXiv  Detail & Related papers  (2023-07-14T18:39:04Z)
• Accelerating exploration and representation learning with offline pre-training [52.6912479800592]
  We show that exploration and representation learning can be improved by separately learning two different models from a single offline dataset. We show that learning a state representation using noise-contrastive estimation and a model of auxiliary reward can significantly improve the sample efficiency on the challenging NetHack benchmark.
  arXiv  Detail & Related papers  (2023-03-31T18:03:30Z)
• Task Formulation Matters When Learning Continually: A Case Study in Visual Question Answering [58.82325933356066]
  Continual learning aims to train a model incrementally on a sequence of tasks without forgetting previous knowledge. We present a detailed study of how different settings affect performance for Visual Question Answering.
  arXiv  Detail & Related papers  (2022-09-30T19:12:58Z)
• Self-Supervised Visual Representation Learning Using Lightweight Architectures [0.0]
  In self-supervised learning, a model is trained to solve a pretext task, using a data set whose annotations are created by a machine. We critically examine the most notable pretext tasks to extract features from image data. We study the performance of various self-supervised techniques keeping all other parameters uniform.
  arXiv  Detail & Related papers  (2021-10-21T14:13:10Z)
• Distill on the Go: Online knowledge distillation in self-supervised learning [1.1470070927586016]
  Recent works have shown that wider and deeper models benefit more from self-supervised learning than smaller models. We propose Distill-on-the-Go (DoGo), a self-supervised learning paradigm using single-stage online knowledge distillation. Our results show significant performance gain in the presence of noisy and limited labels.
  arXiv  Detail & Related papers  (2021-04-20T09:59:23Z)
• Towards Understanding Ensemble, Knowledge Distillation and Self-Distillation in Deep Learning [93.18238573921629]
  We study how Ensemble of deep learning models can improve test accuracy, and how the superior performance of ensemble can be distilled into a single model. We show that ensemble/knowledge distillation in deep learning works very differently from traditional learning theory. We prove that self-distillation can also be viewed as implicitly combining ensemble and knowledge distillation to improve test accuracy.
  arXiv  Detail & Related papers  (2020-12-17T18:34:45Z)

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.