Self-Distillation of Hidden Layers for Self-Supervised Representation Learning

• URL: http://arxiv.org/abs/2603.15553v1
• Date: Mon, 16 Mar 2026 17:13:27 GMT
• Title: Self-Distillation of Hidden Layers for Self-Supervised Representation Learning
• Abstract summary: We introduce Bootleg, a method that bridges the divide by tasking the model with predicting latent representations from hidden layers of a teacher network.<n>We demonstrate that Bootleg significantly outperforms comparable baselines on classification of ImageNet-1K and iNaturalist-21, and semantic segmentation of ADE20K and Cityscapes.
• Score: 18.042078225633045
• License: http://creativecommons.org/licenses/by-sa/4.0/
• Abstract: The landscape of self-supervised learning (SSL) is currently dominated by generative approaches (e.g., MAE) that reconstruct raw low-level data, and predictive approaches (e.g., I-JEPA) that predict high-level abstract embeddings. While generative methods provide strong grounding, they are computationally inefficient for high-redundancy modalities like imagery, and their training objective does not prioritize learning high-level, conceptual features. Conversely, predictive methods often suffer from training instability due to their reliance on the non-stationary targets of final-layer self-distillation. We introduce Bootleg, a method that bridges this divide by tasking the model with predicting latent representations from multiple hidden layers of a teacher network. This hierarchical objective forces the model to capture features at varying levels of abstraction simultaneously. We demonstrate that Bootleg significantly outperforms comparable baselines (+10% over I-JEPA) on classification of ImageNet-1K and iNaturalist-21, and semantic segmentation of ADE20K and Cityscapes.

Related papers

• Gradient Concentration, Not Weight Saliency, Explains Representation-Level Class Unlearning [3.33554314544636]
  We perform the first controlled ablation of the saliency masking mechanism used by SalUn.<n>We find that representation-level forgetting is governed by gradient concentration and representation geometry rather than by the specific identity of saliency-selected weights.
  arXiv  Detail & Related papers  (2026-07-23T14:17:51Z)
• HCL-FF: Hierarchical and Contrastive Learning for Forward-Forward Algorithm [23.026010149504387]
  We propose a hierarchical and contrastive learning framework (HCL-FF) for deep neural networks.<n>HCL-FF introduces a coarse-to-fine hierarchical learning strategy that guides representations from low-level cues to high-level semantics.<n>Experiments on CIFAR-10, CIFAR-100, and Tiny-ImageNet demonstrate that HCL-FF achieves new state-of-the-art performance among FF-based methods.
  arXiv  Detail & Related papers  (2026-05-24T00:48:13Z)
• TD-JEPA: Latent-predictive Representations for Zero-Shot Reinforcement Learning [63.73629127832652]
  We introduce TD-JEPA, which leverages TD-based latent-predictive representations into unsupervised RL.<n> TD-JEPA trains explicit state and task encoders, a policy-conditioned multi-step predictor, and a set of parameterized policies directly in latent space.<n> Empirically, TD-JEPA matches or outperforms state-of-the-art baselines on locomotion, navigation, and manipulation tasks across 13 datasets.
  arXiv  Detail & Related papers  (2025-10-01T10:21:18Z)
• Targeted Forgetting of Image Subgroups in CLIP Models [30.78624907082701]
  Foundation models (FMs) such as CLIP have demonstrated impressive zero-shot performance across various tasks.<n>They often inherit harmful or unwanted knowledge from noisy internet-sourced datasets.<n>Existing model unlearning methods either rely on access to pre-trained datasets or focus on coarse-grained unlearning.<n>We propose a novel three-stage approach that progressively unlearns targeted knowledge while mitigating over-forgetting.
  arXiv  Detail & Related papers  (2025-06-03T17:50:03Z)
• Bayesian Learning-driven Prototypical Contrastive Loss for Class-Incremental Learning [42.14439854721613]
  This paper proposes a method to learn an effective representation between previous and newly encountered class prototypes.<n>We introduce a contrastive loss that incorporates novel classes into the latent representation by reducing intra-class and increasing inter-class distance.
  arXiv  Detail & Related papers  (2024-05-17T19:49:02Z)
• 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)
• Skeleton2vec: A Self-supervised Learning Framework with Contextualized Target Representations for Skeleton Sequence [56.092059713922744]
  We show that using high-level contextualized features as prediction targets can achieve superior performance. Specifically, we propose Skeleton2vec, a simple and efficient self-supervised 3D action representation learning framework. Our proposed Skeleton2vec outperforms previous methods and achieves state-of-the-art results.
  arXiv  Detail & Related papers  (2024-01-01T12:08:35Z)
• Unifying Self-Supervised Clustering and Energy-Based Models [9.3176264568834]
  We establish a principled connection between self-supervised learning and generative models.<n>We show that our solution can be integrated into a neuro-symbolic framework to tackle a simple yet non-trivial instantiation of the symbol grounding problem.
  arXiv  Detail & Related papers  (2023-12-30T04:46:16Z)
• Read Between the Layers: Leveraging Multi-Layer Representations for Rehearsal-Free Continual Learning with Pre-Trained Models [15.847302755988506]
  We address the Continual Learning problem, wherein a model must learn a sequence of tasks from non-stationary distributions. We propose LayUP, a new prototype-based approach to CL that leverages second-order feature statistics from multiple intermediate layers of a pre-trained network. Our results demonstrate that fully exhausting the representational capacities of pre-trained models in CL goes well beyond their final embeddings.
  arXiv  Detail & Related papers  (2023-12-13T13:11:44Z)
• From Pretext to Purpose: Batch-Adaptive Self-Supervised Learning [32.18543787821028]
  This paper proposes an adaptive technique of batch fusion for self-supervised contrastive learning. It achieves state-of-the-art performance under equitable comparisons. We suggest that the proposed method may contribute to the advancement of data-driven self-supervised learning research.
  arXiv  Detail & Related papers  (2023-11-16T15:47:49Z)
• RanPAC: Random Projections and Pre-trained Models for Continual Learning [59.07316955610658]
  Continual learning (CL) aims to learn different tasks (such as classification) in a non-stationary data stream without forgetting old ones. We propose a concise and effective approach for CL with pre-trained models.
  arXiv  Detail & Related papers  (2023-07-05T12:49:02Z)
• Unifying Synergies between Self-supervised Learning and Dynamic Computation [53.66628188936682]
  We present a novel perspective on the interplay between SSL and DC paradigms. We show that it is feasible to simultaneously learn a dense and gated sub-network from scratch in a SSL setting. The co-evolution during pre-training of both dense and gated encoder offers a good accuracy-efficiency trade-off.
  arXiv  Detail & Related papers  (2023-01-22T17:12:58Z)
• Contrastive Learning for Fair Representations [50.95604482330149]
  Trained classification models can unintentionally lead to biased representations and predictions. Existing debiasing methods for classification models, such as adversarial training, are often expensive to train and difficult to optimise. We propose a method for mitigating bias by incorporating contrastive learning, in which instances sharing the same class label are encouraged to have similar representations.
  arXiv  Detail & Related papers  (2021-09-22T10:47:51Z)
• Improving Music Performance Assessment with Contrastive Learning [78.8942067357231]
  This study investigates contrastive learning as a potential method to improve existing MPA systems. We introduce a weighted contrastive loss suitable for regression tasks applied to a convolutional neural network. Our results show that contrastive-based methods are able to match and exceed SoTA performance for MPA regression tasks.
  arXiv  Detail & Related papers  (2021-08-03T19:24:25Z)

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.