Skip to contentScholarGate
LibraryBookshelfDeskReview StudioAssistant
Sign in
On this page
IntuitionHow it worksWhen to use itStrengths & limitationsCommon pitfallsApplicationsFrequently asked🔒 Read the full methodSourcesRelated methods
Cite this pageSpotted an issue on this page? Report or suggest a fix →
Home›Deep learning›Knowledge Distillation
Machine learning

Knowledge Distillation

Knowledge Distillation (Teacher–Student Model Compression) · Also known as: Bilgi Damıtma (Knowledge Distillation), bilgi damıtma, teacher-student distillation, model distillation

Knowledge Distillation is a model-compression technique, introduced by Geoffrey Hinton and colleagues in 2015, that trains a small student model using the soft-label outputs of a large teacher model. Distilled models such as DistilBERT and TinyBERT reach roughly 97% of the larger model's performance while running far faster.

ScholarGate
  1. Machine learning
  2. v1
  3. 2 Sources
  4. PUBLISHED
Cite this page →
Tools & resources
Download slides
Learn & explore

Read the full method

Members only

Sign in with a free account to read this section.

Sign in

Method map

The neighbourhood of related methods — select a node to explore.

Knowledge Distillation
Longformer / BigBirdMixture of ExpertsRandom ForestVisual Contrastive Learn…XGBoostCapsule NetworkEnsemble Self-supervised…Federated LearningMobileNetMultitask Learning

+3 more

When to use it

Use it when you have a strong but heavy trained teacher model and want a smaller, faster student for deployment on text or continuous-feature tasks, with at least about 100 task-specific observations. It assumes the teacher model is already trained and available and that a task-specific dataset is ready. Below roughly 500 observations the student cannot distil enough from the teacher and overfitting becomes a risk; below about 100, distillation is pointless and classic machine learning is the safer choice.

Strengths & limitations

Strengths
  • Yields a much smaller, faster model that can recover around 97% of the teacher's performance.
  • The teacher's soft labels carry richer information than hard labels, improving the student's generalisation.
  • Assumption-light: does not require normally distributed data.
  • Reuses an existing trained teacher, so it suits deployment and serving where speed and size matter.
Limitations
  • Requires an already-trained, available teacher model — distillation cannot start without one.
  • Needs a task-specific dataset; with too little data the student learns little from the teacher.
  • On small datasets (n below about 500) the student overfits and fails to absorb the teacher's knowledge.
  • Adds a training pipeline and tuning burden (temperature, mixing weight) on top of the original model.

Frequently asked

Why use soft labels instead of just the correct answer?

The teacher's full probability distribution reveals how similar the classes look to it and how confident it is. This extra signal — sometimes called dark knowledge — helps the student generalise better than hard labels alone.

How much smaller can the student be?

Distilled models such as DistilBERT and TinyBERT are substantially smaller and faster than their teachers while reaching roughly 97% of the teacher's performance, though the exact trade-off depends on the task and architecture.

What do I need before distilling?

You need a teacher model that is already trained and available, plus a task-specific dataset. With fewer than about 500 observations the student struggles to distil enough from the teacher and may overfit.

What is the alpha in the loss?

Alpha mixes the two parts of the distillation loss: a cross-entropy term on the true labels and a Kullback–Leibler divergence term that aligns the student's outputs with the teacher's. Tuning it balances learning from ground truth versus learning from the teacher.

Sources

  1. Hinton, G., Vinyals, O. & Dean, J. (2015). Distilling the Knowledge in a Neural Network. NeurIPS Deep Learning Workshop. link ↗
  2. Sanh, V., Debut, L., Chaumond, J. & Wolf, T. (2019). DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter. arXiv:1910.01108. link ↗

How to cite this page

ScholarGate. (2026, June 1). Knowledge Distillation (Teacher–Student Model Compression). ScholarGate. https://scholargate.app/en/deep-learning/knowledge-distillation

Related methods

Longformer / BigBirdMixture of ExpertsRandom ForestVisual Contrastive LearningXGBoost

Which method?

Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.

  • Longformer / BigBirdDeep learning↔ compare
  • Mixture of ExpertsDeep learning↔ compare
  • Random ForestMachine learning↔ compare
  • Visual Contrastive LearningDeep learning↔ compare
  • XGBoostMachine learning↔ compare
Compare side by side →

Referenced by

Capsule NetworkEnsemble Self-supervised LearningFederated LearningMobileNetMultitask LearningNeural Architecture SearchSelf-supervised Image ClassificationWeakly supervised vision transformer

Similar methods

BERT Fine-TuningTransfer LearningFine-Tuned TransformerTransfer Learning with BERT-based ClassificationSelf-supervised BERT-based classificationSemi-supervised TransformerWeakly supervised transformerSelf-supervised Transformer

Related reference concepts

Self-Supervised and Representation LearningUnsupervised LearningSupervised LearningDimensionality ReductionVC Dimension and CapacityVariational Inference

Spotted an issue on this page? Report or suggest a fix →

ScholarGate — Knowledge Distillation (Knowledge Distillation (Teacher–Student Model Compression)). Retrieved 2026-07-21 from https://scholargate.app/en/deep-learning/knowledge-distillation · Dataset: https://doi.org/10.5281/zenodo.20539026
Quick facts
Originator
Hinton, G., Vinyals, O. & Dean, J.
Year
2015
Type
Neural network compression (teacher–student)
Task
Classification & prediction
MinSample
100
Related methods
Longformer / BigBirdMixture of ExpertsRandom ForestVisual Contrastive LearningXGBoost
ScholarGate

A content-first reference library for research methods — what each one is, how it works, and where it comes from.

Open data (CC-BY)

Explore

  • Library
  • Search the library…
  • Browse by field
  • Fields
  • Journey
  • Compare
  • Which method?

Reference

  • Subjects
  • Atlas
  • Glossary
  • Methodology
  • Philosophy

Your tools

  • Bookshelf
  • Desk
  • Chat

Company

  • About
  • Pricing
  • Contact
  • Suggest a method

Entries are compiled from published sources for reference. Verifying the accuracy and suitability of any information for your own use remains your responsibility.

© 2026 ScholarGate · A research-method reference library
  • Privacy
  • Cookies
  • Terms
  • Delete account