AI Paper Atlas

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

Entropy-Monitored Kernelized Token Distillation for Audio-Visual Compression

Hyoungseob Park, Lipeng Ke, Pritish Mohapatra, Huajun Ying, sankar venkataraman, Alex Wong

ICLR 2026 · Label-efficient and robust learning › Knowledge distillation · approach: Few-shot and metric distillation

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 kernelized token distillation (KTD) to KDViT-Tiny student · VGGSoundaccuracy (%): 56.1 -> 61.4Improved
add entropy-monitored modality regularization to KTD + KD (ablation)ViT-Tiny student · VGGSoundaccuracy (%): 61.4 -> 62.0Improved
add entropy monitor to Hinton KD (ablation)ViT-Tiny student · VGGSoundaccuracy (%): 56.1 -> 59.9Improved
use RBF kernel instead of linear kernel (ablation)KTD · VGGSoundaccuracy (%): 60.2 -> 61.4Improved
use RBF kernel (gamma 0.5) instead of order-2 polynomial (ablation)KTD · VGGSoundaccuracy (%): 60.5 -> 61.4Improved
use KTD with linear kernel instead of MTST (softmax + masking)KD-based distillation · VGGSoundaccuracy (%): 57.6 -> 60.2Improved
distill the last layer instead of the first layer (ablation)KTD · VGGSoundaccuracy (%): 56.4 -> 60.2Improved
use instance-wise KTD instead of batch-wise (ablation)KTD · VGGSoundaccuracy (%): 59.7 -> 61.2Improved
add cross-modal kernelized token loss (ablation)KTD · VGGSoundaccuracy (%): 61.2 -> 61.0Hurt
distill all modalities instead of vision tokens only (ablation)KTD + KD · VGGSoundaccuracy (%): 54.5 -> 61.2Improved
use 1-layer linear entropy monitor instead of 3-layer MLP (ablation)EM-KTD + KD · VGGSoundaccuracy (%): 61.7 -> 62.0No clear effect
use sliding window kernelization (ablation)EM-KTD + KD · VGGSoundaccuracy (%): 61.9 -> 62.0No 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