AI Paper Atlas

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

MixKD: Towards Efficient Distillation of Large-scale Language Models

Kevin J Liang, Weituo Hao, Dinghan Shen, Yufan Zhou, Weizhu Chen, Changyou Chen, Lawrence Carin

ICLR 2021 · Label-efficient and robust learning › Knowledge distillation · approach: Rethinking distillation objectives · 29 citations (OpenAlex)

This paper tried a common change: Add mixup or CutMix

Add MixKD (student mixup, teacher mixup distillation, and backtranslation) · GLUE test tasks · accuracy or F1 Mixed

What 547 accepted papers report for this change

See this change in detailMatched by the wording of each paper's own result, read by AI from the full text. Papers report gains more often than failures, so this shows what gets published, not the chance a change works.

Read the paper

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 MixKD (student mixup, teacher mixup distillation, and backtranslation)BERT6 student · GLUE test tasksaccuracy or F1Mixed
Add student mixup loss to teacher mixup distillation with backtranslation (ablation)BERT6-TMKD+BT · MRPC devF1: 89.79±0.27 -> 89.64±0.38Hurt
Add student mixup loss to teacher mixup distillation with backtranslation (ablation)BERT6-TMKD+BT · MNLI-m devaccuracy: 82.05±0.11 -> 82.41±0.12Improved
Add student mixup loss to teacher mixup distillation with backtranslation (ablation)BERT6-TMKD+BT · QNLI devaccuracy: 88.42±0.06 -> 88.76±0.15Improved
Add student mixup loss to teacher mixup distillation with backtranslation (ablation)BERT6-TMKD+BT · RTE devaccuracy: 69.37±0.50 -> 68.02±0.11Hurt
Increase mixup ratio above 1 (ablation)MixKD · SST-2accuracyNo clear effect
Add MixKD rather than the TinyBERT data augmentation moduleBERT6 · MNLIaccuracy: 81.5 -> 82.5Improved
Add MixKD rather than the TinyBERT data augmentation moduleBERT6 · SST-2accuracy: 91.3 -> 92.1Improved
Use MixKD instead of fine-tuning alone with limited training dataBERT3 · QQP, MNLI-m, MNLI-mm, and QNLI at 10% and 1% dataaccuracy: +2.0%, +3.0%; +3.9%, +3.4%; +4.4%, +3.3%; +2.4%, +4.1%Improved

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