paper-with-me

홈 › Papers

A Closer Look at Knowledge Distillation with Features, Logits, and Gradients

2022-03-18 · Yen-Chang Hsu, James Smith, Yilin Shen, Zsolt Kira, Hongxia Jin

Knowledge distillation (KD) is a substantial strategy for transferring learned knowledge from one neural network model to another. A vast number of methods have been developed for this strategy. While most method designs a more efficient way to facilitate knowledge transfer, less attention has been put on comparing the effect of knowledge sources such as features, logits, and gradients. This work provides a new perspective to motivate a set of knowledge distillation strategies by approximating the classical KL-divergence criteria with different knowledge sources, making a systematic comparison possible in model compression and incremental learning. Our analysis indicates that logits are generally a more efficient knowledge source and suggests that having sufficient feature dimensions is crucial for the model design, providing a practical guideline for effective KD-based transfer learning.

📄 PDF Abstract BibTeX arXiv:2203.10163

Code (0)

등록된 구현이 없습니다.

Tasks

Incremental LearningKnowledge DistillationModel CompressionTransfer Learning

Methods 이 논문이 사용한 방법론

Knowledge Distillation A very simple way to improve the performance of almost any machine learning algorithm is to train many different models on the same data and then to average their predictions.…

Similar Papers 제목 키워드 기반

A Closer Look at Knowledge Distillation in Spiking Neural Network Training

2025-11-10 · Xu Liu, Na Xia, Jinxing Zhou, Jingyuan Xu 외 arxiv

Spiking Neural Networks (SNNs) become popular due to excellent energy efficiency, yet facing challenges for effective model training. Recent works improve this by introducing knowledge distillation (KD) techniques, with …

Knowledge Distillation

NormKD: Normalized Logits for Knowledge Distillation

2023-08-01 · Zhihao Chi, Tu Zheng, Hengjia Li, Zheng Yang 외

Logit based knowledge distillation gets less attention in recent years since feature based methods perform better in most cases. Nevertheless, we find it still has untapped potential when we re-investigate the temperatur…

image-classificationImage ClassificationKnowledge Distillation

Heterogeneous Complementary Distillation

2025-11-14 · Liuchi Xu, Hao Zheng, Lu Wang, Lisheng Xu 외 arxiv

Knowledge distillation (KD)transfers the dark knowledge from a complex teacher to a compact student. However, heterogeneous architecture distillation, such as Vision Transformer (ViT) to ResNet18, faces challenges due to…

Knowledge Distillation

TopKD: Top-scaled Knowledge Distillation

2025-08-06 · Qi Wang, Jinjia Zhou arxiv

Recent advances in knowledge distillation (KD) predominantly emphasize feature-level knowledge transfer, frequently overlooking critical information embedded within the teacher's logit distributions. In this paper, we re…

Knowledge Distillation

Decoupling Dark Knowledge via Block-wise Logit Distillation for Feature-level Alignment

2024-11-03 · Chengting Yu, Fengzhao Zhang, Ruizhe Chen, Aili Wang 외

Knowledge Distillation (KD), a learning manner with a larger teacher network guiding a smaller student network, transfers dark knowledge from the teacher to the student via logits or intermediate features, with the aim o…

Knowledge DistillationPhilosophy