Papers › Information Theoretic Representation Distillation

Information Theoretic Representation Distillation

1 Dec 2021arXiv:2112.00459archive 2025-07-28

Roy Miles, Adrian Lopez Rodriguez, Krystian Mikolajczyk

Despite the empirical success of knowledge distillation, current state-of-the-art methods are computationally expensive to train, which makes them difficult to adopt in practice. To address this problem, we introduce two distinct complementary losses inspired by a cheap entropy-like estimator. These losses aim to maximise the correlation and mutual information between the student and teacher representations. Our method incurs significantly less training overheads than other approaches and achieves competitive performance to the state-of-the-art on the knowledge distillation and cross-model transfer tasks. We further demonstrate the effectiveness of our method on a binary distillation task, whereby it leads to a new state-of-the-art for binary quantisation and approaches the performance of a full precision model. Code: www.github.com/roymiles/ITRD

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

roymiles/ITRD officialmentioned in papermentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Classification with Binary Weight NetworkKnowledge DistillationQuestion Answering

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Knowledge Distillation CIFAR-100 resnet8x4 (T: resnet32x4 S: resnet8x4) Top-1 Accuracy (%) 76.68 #9 of 27 Archive leaderboard report
Knowledge Distillation CIFAR-100 vgg8 (T:vgg13 S:vgg8) Top-1 Accuracy (%) 74.93 #16 of 27 Archive leaderboard report
Knowledge Distillation CIFAR-100 resnet110 (T:resnet110 S:resnet20) Top-1 Accuracy (%) 71.99 #23 of 27 Archive leaderboard report
Knowledge Distillation ImageNet ITRD (T: ResNet-34 S:ResNet-18) CRD training setting ✓ #41 of 52 Archive leaderboard report
Knowledge Distillation ImageNet ITRD (T: ResNet-34 S:ResNet-18) Top-1 accuracy % 71.68 #41 of 52 Archive leaderboard report
Knowledge Distillation ImageNet ITRD (T: ResNet-34 S:ResNet-18) model size 11.69M #41 of 52 Archive leaderboard report
Question Answering SQuAD1.1 BERT - 6 Layers EM 81.5 #50 of 213 Archive leaderboard report
Question Answering SQuAD1.1 BERT - 6 Layers F1 88.5 #50 of 213 Archive leaderboard report
Question Answering SQuAD1.1 BERT - 3 Layers EM 77.7 #97 of 213 Archive leaderboard report
Question Answering SQuAD1.1 BERT - 3 Layers F1 85.8 #97 of 213 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Knowledge Distillation

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections