KDM unifies feature matching in neural networks.
problem Matching neural network activations for knowledge transfer.
method Formalizes distribution matching strategy.
result New theoretical results for KD.
On-device research index
arXiv research
A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.
168,742 papers · 148 categories
Trend · papers per month
48 results for “knowledge distillation (KD)”
This study improves knowledge distillation for RNN-T models with noisy labels.
problem Challenges in distilling knowledge from RNN-T models with variable quality teachers.
method Full-sum distillation and sequence-level knowledge distillation.
result Full-sum distillation outperforms other methods for RNN-T models, especially for bad teachers.
Knowledge distillation (KD) is a technique to derive optimal performance from a small student network (SN) by distilling knowledge of a large teacher network (TN) and transferring the distilled knowledge to the small SN. Since a role of convolutional neural network (CNN) in KD is to embed a dataset so as to perform a g…
Unified framework for image classification and regression using cGAN-generated samples.
problem Lack of unified KD methods for both classification and regression tasks.
method cGAN-KD framework based on cGANs.
result Unified framework for both classification and regression tasks, compatible with other KD methods.
Pea-KD improves BERT student models by 4.4% on average in GLUE tasks.
problem Efficiently compressing BERT while maintaining performance.
method Parameter-efficient and accurate Knowledge Distillation (Pea-KD) with Shuffled Parameter Sharing (SPS) and Pretraining with Teacher's Predictions (PTP).
result Pea-KD improves student model's performance by 4.4% on average in four GLUE tasks.
PS-KD distills a model's own knowledge to soften hard targets during training.
problem Improving generalization of deep neural networks by softening hard targets.
method Progressive self-knowledge distillation (PS-KD) that progressively distills a model's own knowledge to soften hard targets.
result PS-KD improves accuracy and provides high quality of confidence estimates in terms of calibration and ordinal ranking.
Study explores various KD methods to improve model performance.
problem Improving model performance with limited data.
method Nine different KD methods covering various aspects of the teacher.
result KD framework provides generalization gains over standard training.
This paper improves understanding of knowledge distillation and its effects.
problem Improving model quality with a fixed capacity budget.
method Categorizing teacher's knowledge into three levels and studying their effects on distillation.
result The three hierarchical levels of teacher's knowledge (universe, domain, instance) significantly impact knowledge distillation.
Proposes LsrKD and MrKD to improve neural network training performance.
problem Improving neural network training performance, especially on deep networks.
method Extends Label Smoothing Regularization with Self-Knowledge Distillation, introducing LsrKD and MrKD.
result LsrKD and MrKD significantly improve training performance on deep neural networks.
Role-wise data augmentation improves knowledge distillation effectiveness.
problem Existing knowledge distillation methods fail to utilize the full potential of teacher-student data interaction.
method Design and implement data augmentation agents with distinct roles for teacher and student.
result Specially tailored data points enhance the demonstration of teacher's knowledge to the student.
Survey examines distillation methods for large language models.
problem Efficiently compress large language models while preserving their capabilities.
method Knowledge Distillation and Dataset Distillation techniques.
result Integrating KD and DD can produce more effective and scalable compression strategies.
New federated learning protocols improve on knowledge distillation's poor performance.
problem Designing a universal API for federated learning without public data.
method Proposed Federated Kernel ridge regression using knowledge distillation.
result Performance of new protocols closely matches theoretical predictions.
Zero-shot KD for object detection without training data.
problem Challenges in using training data for knowledge distillation.
method Synthesizes pseudo-targets and samples using pretrained network.
result Achieves respectable mAP on object detection benchmarks.
Current approaches for Knowledge Distillation (KD) either directly use training data or sample from the training data distribution. In this paper, we demonstrate effectiveness of 'mismatched' unlabeled stimulus to perform KD for image classification networks. For illustration, we consider scenarios where this is a comp…
Knowledge distillation (KD) is a very popular method for model size reduction. Recently, the technique is exploited for quantized deep neural networks (QDNNs) training as a way to restore the performance sacrificed by word-length reduction. KD, however, employs additional hyper-parameters, such as temperature, coeffici…
A new KD layer lets student models learn and apply teacher knowledge explicitly.
problem Implicit action of traditional KD on student's feature transform limits its use in intermediate layers.
method Proposes a learnable KD layer that explicitly embeds teacher's knowledge in feature transform.
result Improves KD with two abilities: leveraging teacher's knowledge and feeding forward knowledge deeper.
SPEQ improves quantized neural networks by stochastic precision sharing and cosine similarity loss.
problem Improving quantized deep neural networks for edge devices.
method SPEQ combines stochastic precision sharing and cosine similarity loss for knowledge distillation.
result SPEQ outperforms existing methods in various tasks.
Policy-Oriented Binary Classification: Improving (KD-)CART Final Splits for Subpopulation Targetingstat.ML
New split rules improve subpopulation targeting in policy-making.
problem Improving binary classification for subpopulation targeting in policy-making.
method MDFS, PFS, wEFS for maximizing distance and penalizing final splits.
result Proposed methods target more vulnerable subpopulations than classic CART/KD-CART.
Born Again Neural Networksstat.ML
Knowledge Distillation (KD) consists of transferring âknowledgeâ from one machine learning model (the teacher) to another (the student). Commonly, the teacher is a high-capacity model with formidable performance, while the student is more compact. By transferring knowledge, one hopes to benefit from the studentâs…
Paper proposes a data-free adversarial distillation method.
problem Tackles the challenge of model compression without real-world data.
method Introduces a novel adversarial distillation mechanism to craft a compact student model without any real-world data.
result Demonstrates comparable performance to data-driven methods and achieves state-of-the-art results in semantic segmentation.
KD can lead to student-teacher deviations that improve performance.
problem KD can lead to student-teacher deviations that may outperform the teacher.
method Characterized and explained the nature of student-teacher deviations through experiments and theory.
result KD can lead to improved generalization by exaggerating the implicit bias of gradient descent.
Knowledge distillation (KD) is a well-known method to reduce inference latency by compressing a cumbersome teacher model to a small student model. Despite the success of KD in the classification task, applying KD to recommender models is challenging due to the sparsity of positive feedback, the ambiguity of missing fee…
RKD improves model compression by distilling residual knowledge from a deep teacher model.
problem Performance degradation due to the gap between student and teacher models.
method Introduces an assistant model to distill residual knowledge from the teacher model.
result RKD achieves better results on popular classification datasets than state-of-the-art methods.
KD-Net transfers knowledge from multi-modal to mono-modal segmentation networks.
problem Limited acquisition of multiple imaging modalities in clinical settings.
method Generalized distillation framework adapted for mono-modal networks.
result The student network outperforms baseline mono-modal networks in brain tumor segmentation.
This paper improves ASR performance by aligning frames more accurately.
problem Disagreement between teacher-student models in frame-level alignment.
method Introduces self-knowledge distillation (SKD) to guide frame-level alignment.
result Improves both resource efficiency and performance in ASR.
The holy grail in deep neural network research is porting the memory- and computation-intensive network models on embedded platforms with a minimal compromise in model accuracy. To this end, we propose a novel approach, termed as Variational Student, where we reap the benefits of compressibility of the knowledge distil…
A new distillation method transfers channel information from teacher to student.
problem Transfer knowledge from teacher to student with fewer parameters and calculations.
method Channel Distillation (CD) and Guided Knowledge Distillation (GKD) with loss decay.
result Achieved 27.68% top-1 error on ImageNet with ResNet18, outperforming state-of-the-art methods.
New method distills cloud models into edge-friendly ones.
problem Cloud-to-edge model compression with limited data exchange.
method Two-step workflow of deprivatization and distillation.
result Outperforms previous state-of-the-art approaches on various benchmarks.
In many situations, we need to build and deploy separate models in related environments with different data qualities. For example, an environment with strong observation equipments (e.g., intensive care units) often provides high-quality multi-modal data, which are acquired from multiple sensory devices and have rich-…
New method observes learning paths to improve model supervision.
problem Improving model performance through better supervision.
method Observing learning paths to refine labels and propose Filter-KD.
result Models can refine bad labels through a 'zig-zag' learning path.
CoDistill-GRPO improves small models in GRPO by distilling knowledge from a larger model.
problem Small models in GRPO struggle with sparse rewards on difficult tasks.
method Simultaneously trains a large and small model using co-distillation and GRPO objectives.
result Significant improvement in small model performance over standard GRPO on mathematical benchmarks.
MixKD improves large-scale language model compression and generalization.
problem Inefficient and resource-intensive large-scale language models.
method MixKD uses mixup data augmentation to enhance student model's generalization ability.
result MixKD leads to significant performance gains over standard KD and competitive baselines.
Ranking Distillation: Learning Compact Ranking Models With High Performance for Recommender Systemcs.LG
We propose a novel way to train ranking models, such as recommender systems, that are both effective and efficient. Knowledge distillation (KD) was shown to be successful in image recognition to achieve both effectiveness and efficiency. We propose a KD technique for learning to rank problems, called \emph{ranking dist…
A method for faster neural architecture search using low-fidelity training.
problem Time-consuming evaluations in neural architecture search.
method Bayesian multi-fidelity method with knowledge distillation.
result Training for a few epochs with knowledge distillation leads to better architecture selection.
Knowledge distillation (KD) is a popular method for reducing the computational overhead of deep network inference, in which the output of a teacher model is used to train a smaller, faster student model. Hint training (i.e., FitNets) extends KD by regressing a student model's intermediate representation to a teacher mo…
In the absence of sufficient data variation (e.g., scanner and protocol variability) in annotated data, deep neural networks (DNNs) tend to overfit during training. As a result, their performance is significantly lower on data from unseen sources compared to the performance on data from the same source as the training …
This paper introduces GLT for better input data representation in BNN and proposes a compact topology with block pruning.
problem Improving input data representation for Binary Neural Networks (BNN).
method Generic Learned Thermometer (GLT) for encoding, block pruning and Knowledge Distillation for compact topology.
result Significant accuracy gains and lightweight fully-binarized models with limited accuracy degradation.
Recent variants improve knowledge distillation performance.
problem Improving the performance of knowledge distillation.
method Introducing additional components or changing the learning process.
result These variants have shown promising results.
MAZE attacks a model without data, using synthetic inputs.
problem Model Stealing attacks that require data to replicate a target model.
method MAZE uses zeroth-order gradient estimation to create synthetic inputs.
result MAZE achieves high clone accuracy without data.
This paper proposes a method to compress and adapt CNNs for real-world applications.
problem Differences in data distributions and high computational costs limit CNN adoption.
method Joint optimization of CNNs for unsupervised domain adaptation and knowledge distillation.
result The proposed method achieves the highest accuracy with comparable or lower time complexity.
Survey of knowledge distillation for resource-limited devices.
problem Deploying large deep learning models on resource-limited devices.
method Knowledge distillation using a smaller model trained with information from a larger model.
result A new metric (distillation metric) for comparing different knowledge distillation algorithms.
RID framework quantifies and regularizes task-relevant knowledge in distillation.
problem Distilling irrelevant information can hinder student model performance.
method Partial Information Decomposition to quantify and regularize task-relevant knowledge.
result RID framework leads to more resilient distillation under nuisance teachers.
The paper explains knowledge distillation by analyzing visual concepts in DNNs.
problem Understanding how knowledge distillation affects the learning of visual concepts in deep neural networks.
method The paper proposes three hypotheses and designs mathematical metrics to evaluate feature representations of DNNs.
result The hypotheses were verified through experiments on various DNNs.
CAKD framework optimizes knowledge transfer by focusing on influential components of distillation.
problem Balancing and optimizing knowledge transfer in distillation models.
method Decouple KL divergence into BCD, SCD, and WCD; prioritize influential components.
result CAKD framework consistently outperforms baseline across diverse models and datasets.
Often we wish to transfer representational knowledge from one neural network to another. Examples include distilling a large network into a smaller one, transferring knowledge from one sensory modality to a second, or ensembling a collection of models into a single estimator. Knowledge distillation, the standard approa…
Knowledge distillation improves model accuracy by mimicking teacher model probabilities.
problem Improving model accuracy through model compression.
method Casting knowledge distillation as a semiparametric inference problem, deriving new guarantees, and developing enhancements.
result Enhancements improve student performance by mitigating teacher overfitting and underfitting.
Study on ensemble, distillation, and self-distillation in deep learning models.
problem Improving test accuracy in deep learning models using ensemble and distillation methods.
method Formal study of ensemble and distillation, considering multi-view data structure.
result Proven that ensemble and distillation can improve test accuracy in deep learning models, and the superior performance can be distilled into a single model.
Survey of knowledge distillation for model compression.
problem Deploying large deep models on resource-limited devices.
method Knowledge distillation from large teacher models to small student models.
result Effective model compression and acceleration.