AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

Study of Training Dynamics for Memory-Constrained Fine-Tuning

Aël Quélennec, Nour Hezbri, Pavlo Mozharovskyi, Van-Tam Nguyen, Enzo Tartaglione

ICLR 2026 · Architectures and efficiency › Pruning and efficient tuning · approach: Low-rank compression and merging · 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
resample channels dynamically between epochs instead of static selection (ablation)channel selection strategies · CNN fine-tuning on 7 datasetstop-1 accuracyImproved
restrict random channel sampling to top-K layers by RGN (ablation)D-Full Random · MbV2-w0.35, budget 27 946, 7-dataset averagetop-1 accuracy: 74.24 -> 74.91Improved
restrict random channel sampling to top-K layers by RGN (ablation)D-Full Random · MbV2-w0.35, budget 66 592, 7-dataset averagetop-1 accuracy: 75.09 -> 75.62Improved
random channel sampling in top-K layers instead of deterministic highest-RGN selection (ablation)TraDy · CNNs across datasetstop-1 accuracy: 74.49 -> 74.91 (MbV2-w0.35, 27 946)Improved
use TraDy instead of Sparse Update (SU)MbV2-w0.35 · 7-dataset average, budget 27 946top-1 accuracy: 74.22 -> 74.91Improved
use TraDy instead of Velocity on larger budgetsMbV2-w0.35 · MbV2-w0.35, 7-dataset average, budget 66 592top-1 accuracy: 75.73 vs 75.62 (budget 66 592)Hurt
vary number K of top layers for random sampling (ablation)TraDy on MobileNetV2 CIFAR-10 · smallest budgettop-1 accuracyMixed
use reweighted gradient norm instead of raw gradient norm for channel thresholding (ablation)gradient-threshold freezing · MobileNetV2 CIFAR-10accuracy vs memoryImproved
channel-level selection instead of neuron-level (Velocity)activation sparsity · MobileNetV2 Foodactivation sparsity: 20-40% -> 97-99%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