AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

A Mechanism-Driven Theory of Phase Transitions in Active Learning

Julia Machnio, Mads Nielsen, MOSTAFA MEHDIPOUR GHAZI

ECCV 2026 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Invariance-based self-supervised learning · 1 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
use a self-supervised SimCLR backbone instead of supervised trainingsegmented regression phase detection · CIFAR-100second transition tau2 (labels): 37,900 -> 11,000Improved
increase number of segments from K=2 to K=3 (ablation)segmented regression · CIFAR-10BIC: -42188 -> -43235Improved
increase number of segments from K=3 to K=5 (ablation)segmented regression · CIFAR-10SSE: 0.09 -> 0.06Improved
increase segments from K=1 to K=3 (ablation)segmented regression · ISICR^2: 0.828 -> 0.937Improved
increase segments from K=3 to K=4 (ablation)segmented regression · CIFAR-10R^2: 0.998 -> 0.998No clear effect
use TypiClust (representativeness) in the first phaseactive learning strategies · CIFAR-100 (phase I, <=5k labels)accuracy: ~2% accuracy advantageImproved
switch among TypiClust, Coreset, Uncertainty at detected transitions (TCU)static acquisition baselines · CIFAR benchmarksaccuracyImproved
use BYOL embeddings instead of MoCov3 for TypiClustTypiClust · ImageNet-100 low budgetaccuracyHurt
use representativeness or coverage-based selection instead of uncertainty at scaleTypiClust · CIFAR-10 runtimeruntime ratio: 6.79Hurt
use pretrained SSL representations instead of training from scratchactive learning pipeline · training costtraining speedup: 30x to over 160xImproved

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