AI Paper Atlas

Research map / Label-efficient and robust learning / Knowledge distillation

Comprehensive Knowledge Distillation with Causal Intervention

Xiang Deng, Zhongfei Zhang

NeurIPS 2021 · Label-efficient and robust learning › Knowledge distillation · approach: Distillation with causal and adaptive guidance · 14 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add sample representation distillation (ablation)CID · CIFAR-100 WRN-40-2 -> WRN-16-2accuracy (%): +1.51%Improved
add class representation distillation (ablation)CID · CIFAR-100 WRN-40-2 -> WRN-16-2accuracy (%): +0.65%Improved
use causal intervention P(Y|do(X)) instead of P(Y|X) (ablation)CID · CIFAR-100 WRN-40-2 -> WRN-16-2accuracy (%): +0.70%Improved
use normalized MSE (NM_MSE) instead of plain MSE for sample alignment (ablation)CID · CIFAR-100 WRN-40-2 -> WRN-16-2accuracy (%)Improved
distill last-layer features instead of other layers or all layers (ablation)CID · CIFAR-100 WRN-40-2 -> WRN-16-2accuracy (%)Improved
add CID to KD baselineKD · CIFAR-10 WRN-16-4 -> WRN-16-1accuracy (%): 92.55±0.10 -> 92.95±0.10Improved
use CID instead of SRRLSRRL · CIFAR-100 WRN-40-4 -> WRN-16-2accuracy (%): 75.96±0.21 -> 76.40±0.15Improved
use CID instead of CRD on ImageNetCRD · ImageNet top-1accuracy (%): 71.4 -> 71.9Improved
use CID in cross-dataset transferCRD · CIFAR-100 to STL-10accuracy (%): 72.2 -> 72.5Improved
use CID with limited training data (20%)SP · CIFAR-100, 20% dataaccuracy (%): 60.35 -> 62.13Improved
add extra linear layer to competitor methodscompetitor KD methods · CIFAR-100accuracy (%)Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy