AI Paper Atlas

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

Aligning Logits Generatively for Principled Black-Box Knowledge Distillation in the Wild

Xiang Xiang, Jing Ma, Dongrui Wu, Zhigang Zeng, Xilin Chen

TPAMI 2025 · Label-efficient and robust learning › Knowledge distillation · approach: Few-shot and metric distillation · 0 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 information maximization loss (alpha 0.0 -> 0.5), soft responses (ablation)MEKD · CIFAR-100 (ResNet56->MobileNet)top-1 accuracy (%): 60.87 -> 67.07Improved
add information maximization loss (alpha 0.0 -> 0.5), hard responses (ablation)MEKD · CIFAR-100 (ResNet56->MobileNet)top-1 accuracy (%): 61.31 -> 67.36Improved
add KL logit-level loss (beta 0.0 -> 1.0), soft responses (ablation)MEKD · CIFAR-100 (ResNet56->MobileNet)top-1 accuracy (%): 56.28 -> 67.07Improved
add KL logit-level loss (beta 0.0 -> 1.0), hard responses (ablation)MEKD · CIFAR-100 (ResNet56->MobileNet)top-1 accuracy (%): 56.23 -> 67.36Improved
use L1 instead of L2 norm for image-level loss (ablation)MEKD (soft) · CIFAR-10 (ResNet56->MobileNet)top-1 accuracy (%): 86.85 / 86.63No clear effect
use L1 instead of L2 norm for image-level loss (ablation)MEKD (soft) · CIFAR-100 (ResNet56->MobileNet)top-1 accuracy (%): 67.07 / 66.95No clear effect
use hard instead of soft teacher responsesMEKD · Table 1 benchmarkstop-1 accuracy (%): difference of less than 1%No clear effect
increase query data size from 0.1K to 50K (full) on CIFAR-10 (ablation)MEKD (soft) · CIFAR-10 (ResNet56->MobileNet)top-1 accuracy (%): 49.04 -> 93.48Improved
use MEKD instead of DB3KD with only 0.1K samplesMEKD (soft) · CIFAR-10, data size 0.1Ktop-1 accuracy (%): 43.05 -> 49.04Improved
use MEKD instead of DB3KD on ImageNet-1K (RN50 to RN34)MEKD (soft) · ImageNet-1Ktop-1 accuracy (%): 58.61 -> 59.89Improved
use MEKD instead of DB3KD on out-of-domain SVHN (ResNet56 teacher)MEKD (soft) · SVHN (teacher trained on Syn. Digits)top-1 accuracy (%): 90.15 -> 86.45Hurt
add information maximization loss without KL loss (ablation)MEKD · CIFAR-10/100 limited queries (10K to 50K)top-1 accuracy (%)No clear effect

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