What Do Students Learn? A Feature-Level Analysis of Dark Knowledge

• URL: http://arxiv.org/abs/2606.03052v1
• Date: Tue, 02 Jun 2026 02:38:18 GMT
• Title: What Do Students Learn? A Feature-Level Analysis of Dark Knowledge
• Abstract summary: Confusion Distillation (CD) is a teacher-free self-distillation method that utilizes the model's own evolving confusion patterns as dynamic soft targets.<n>CD achieves competitive performance on ResNet-34 and ResNet-50 for CIFAR-100, outperforming existing self-distillation methods like CS-KD and PS-KD by 1.2%.
• Score: 4.060731229044571
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Knowledge Distillation (KD) is a powerful tool for model compression, yet the precise mechanisms by which student models acquire feature representations remain underexplored. In this work, we analyze student feature learning using the Interaction Tensor framework. Our analysis reveals that effective KD acts as a regularizer that prunes low-frequency, sample-specific features, encouraging the student to rely on a compact set of highly reusable features. Crucially, we observe that the dataset-level confusion matrix contains structural information analogous to the teacher's "Dark Knowledge." Leveraging this insight, we propose Confusion Distillation (CD), a teacher-free self-distillation method that utilizes the model's own evolving confusion patterns as dynamic soft targets. CD achieves competitive performance on ResNet-34 and ResNet-50 for CIFAR-100, outperforming existing self-distillation methods like CS-KD and PS-KD by 1.2% while offering a computationally efficient alternative to standard KD.

Related papers

• Learning from Stochastic Teacher Representations Using Student-Guided Knowledge Distillation [64.15918654558816]
  Self-distillation (SSD) training strategy is introduced for filtering and weighting teacher representation to distill from task-relevant representations only.<n> Experimental results on real-world affective computing, wearable/biosignal datasets from the UCR Archive, the HAR dataset, and image classification datasets show that the proposed SSD method can outperform state-of-the-art methods.
  arXiv  Detail & Related papers  (2025-04-19T14:08:56Z)
• Knowledge Distillation: Enhancing Neural Network Compression with Integrated Gradients [0.0]
  This paper proposes a machine learning framework that augments Knowledge Distillation (KD) with Integrated Gradients (IG)<n>We introduce a novel data augmentation strategy where IG maps, precomputed from a teacher model, are overlaid onto training images to guide a compact student model toward critical feature representations.<n>Experiments on CIFAR-10 demonstrate the efficacy of our method: a student model, compressed 4.1-fold from the MobileNet-V2 teacher, achieves 92.5% classification accuracy, surpassing the baseline student's 91.4% and traditional KD approaches, while reducing inference latency from 140 ms to 13 ms--a tenfold
  arXiv  Detail & Related papers  (2025-03-17T10:07:50Z)
• TIE-KD: Teacher-Independent and Explainable Knowledge Distillation for Monocular Depth Estimation [1.03590082373586]
  We introduce a novel Teacher-Independent Explainable Knowledge Distillation (TIE-KD) framework that streamlines the knowledge transfer from complex teacher models to compact student networks. The cornerstone of TIE-KD is the Depth Probability Map (DPM), an explainable feature map that interprets the teacher's output. Extensive evaluation of the KITTI dataset indicates that TIE-KD not only outperforms conventional response-based KD methods but also demonstrates consistent efficacy across diverse teacher and student architectures.
  arXiv  Detail & Related papers  (2024-02-22T07:17:30Z)
• Distilling Privileged Multimodal Information for Expression Recognition using Optimal Transport [46.91791643660991]
  Deep learning models for multimodal expression recognition have reached remarkable performance in controlled laboratory environments. These models struggle in the wild because of the unavailability and quality of modalities used for training. In practice, only a subset of the training-time modalities may be available at test time. Learning with privileged information enables models to exploit data from additional modalities that are only available during training.
  arXiv  Detail & Related papers  (2024-01-27T19:44:15Z)
• Comparative Knowledge Distillation [102.35425896967791]
  Traditional Knowledge Distillation (KD) assumes readily available access to teacher models for frequent inference. We propose Comparative Knowledge Distillation (CKD), which encourages student models to understand the nuanced differences in a teacher model's interpretations of samples. CKD consistently outperforms state of the art data augmentation and KD techniques.
  arXiv  Detail & Related papers  (2023-11-03T21:55:33Z)
• Knowledge Distillation Performs Partial Variance Reduction [93.6365393721122]
  Knowledge distillation is a popular approach for enhancing the performance of ''student'' models. The underlying mechanics behind knowledge distillation (KD) are still not fully understood. We show that KD can be interpreted as a novel type of variance reduction mechanism.
  arXiv  Detail & Related papers  (2023-05-27T21:25:55Z)
• Knowledge Diffusion for Distillation [53.908314960324915]
  The representation gap between teacher and student is an emerging topic in knowledge distillation (KD) We state that the essence of these methods is to discard the noisy information and distill the valuable information in the feature. We propose a novel KD method dubbed DiffKD, to explicitly denoise and match features using diffusion models.
  arXiv  Detail & Related papers  (2023-05-25T04:49:34Z)
• CES-KD: Curriculum-based Expert Selection for Guided Knowledge Distillation [4.182345120164705]
  This paper proposes a new technique called Curriculum Expert Selection for Knowledge Distillation (CES-KD) CES-KD is built upon the hypothesis that a student network should be guided gradually using stratified teaching curriculum. Specifically, our method is a gradual TA-based KD technique that selects a single teacher per input image based on a curriculum driven by the difficulty in classifying the image.
  arXiv  Detail & Related papers  (2022-09-15T21:02:57Z)
• Knowledge Distillation with Representative Teacher Keys Based on Attention Mechanism for Image Classification Model Compression [1.503974529275767]
  knowledge distillation (KD) has been recognized as one of the effective method of model compression to decrease the model parameters. Inspired by attention mechanism, we propose a novel KD method called representative teacher key (RTK) Our proposed RTK can effectively improve the classification accuracy of the state-of-the-art attention-based KD method.
  arXiv  Detail & Related papers  (2022-06-26T05:08:50Z)
• How and When Adversarial Robustness Transfers in Knowledge Distillation? [137.11016173468457]
  This paper studies how and when the adversarial robustness can be transferred from a teacher model to a student model in Knowledge distillation (KD) We show that standard KD training fails to preserve adversarial robustness, and we propose KD with input gradient alignment (KDIGA) for remedy. Under certain assumptions, we prove that the student model using our proposed KDIGA can achieve at least the same certified robustness as the teacher model.
  arXiv  Detail & Related papers  (2021-10-22T21:30:53Z)
• KDExplainer: A Task-oriented Attention Model for Explaining Knowledge Distillation [59.061835562314066]
  We introduce a novel task-oriented attention model, termed as KDExplainer, to shed light on the working mechanism underlying the vanilla KD. We also introduce a portable tool, dubbed as virtual attention module (VAM), that can be seamlessly integrated with various deep neural networks (DNNs) to enhance their performance under KD.
  arXiv  Detail & Related papers  (2021-05-10T08:15:26Z)

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.