mlsanityData to 14 Oct 2026

CAENL: Collapse-Based Active Neural Learning with Objective-Aligned Representation Control

arXiv 2610.17509 ↗First posted 13 Oct 2026cs.LG cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Active learning changes the representations used to acquire subsequent labels. We study this interaction through Collapse-Based Active Neural Learning (CAENL), which separates acquisition, Dynamic Collapse Regularization (DCR), and its coefficient policy. DCR penalizes a normalized within-class to between-class scatter residual above a frozen reference, with a separate class-separation safeguard. Fixed, scheduled, deterministic-feedback, and learned-feedback policies apply this objective. The analysis establishes scatter invariances, conditional local descent, and coefficient bounds, and shows why clean geometry alone cannot guarantee input robustness. Six-seed classification comparisons are complemented by a 104-run study with separate control selection and ten-seed evaluations on each of two backbones. At 60% annotations on ImageNet-100, entropy acquisition with DCR and MACC-Lite reaches 73.420 +/- 0.512% accuracy on ResNet-50 and 73.086 +/- 0.397% on ResNet-18 (mean +/- seed SD), exceeding entropy alone by 1.742 and 1.834 percentage points. Both gains survive Holm correction under paired-t and exact sign-flip tests, with unchanged feedback settings across backbones. Independently tuned constants attain slightly higher mean accuracy; the paired comparisons establish neither adaptive superiority nor equivalence. Separate six-seed spectral studies yield a small fixed-regularization C4 loss reduction, worse CIFAR-10 diffusion FID, and inconclusive AudioCaps captioning effects. Earlier classification studies find near-zero AutoAttack accuracy at L-infinity radius 8/255 and no added benefit from learned control. The results support the evaluated combined classification treatment and identify the task dependence of geometry-informed regularization.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 31 of its 48 references that are in the map. First posted on arXiv 13 Oct 2026.

Builds on

31 of 48 references.

References · 31

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  3. 2019Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformercited by 14,599
  4. 2020Denoising Diffusion Implicit Modelscited by 9,781
  5. 2017On Calibration of Modern Neural Networkscited by 4,960
  6. 2016Improved Techniques for Training GANscited by 2,970
  7. 2020Supervised Contrastive Learningcited by 2,829
  8. 2014CIDEr: Consensus-based Image Description Evaluationcited by 2,046
  9. 2020Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attackscited by 1,451
  10. 2017Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semanticscited by 1,358
  11. 2019Contrastive Multiview Codingcited by 1,301
  12. 2018Demystifying MMD GANscited by 1,293
19 more
  1. 2021VICReg: Variance-Invariance-Covariance Regularization for Self-Supervised Learningcited by 1,057
  2. 2017GradNorm: Gradient Normalization for Adaptive Loss Balancing in Deep Multitask Networkscited by 834
  3. 2016SPICE: Semantic Propositional Image Caption Evaluationcited by 824
  4. 2017Deep Bayesian Active Learning with Image Datacited by 819
  5. 2020Prevalence of Neural Collapse during the terminal phase of deep learning trainingcited by 719
  6. 2019Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Boundscited by 610
  7. 2021A Geometric Analysis of Neural Collapse with Unconstrained Featurescited by 227
  8. 2017Formal Guarantees on the Robustness of a Classifier against Adversarial Manipulationcited by 185
  9. 2016Improved Image Captioning via Policy Gradient optimization of SPIDErcited by 113
  10. 2015Deep Linear Discriminant Analysiscited by 23
  11. 2024Uncertainty Herding: One Active Learning Method for All Label Budgetscited by 15
  12. 2022Deep Active Learning with Noise Stabilitycited by 14
  13. 2024Beyond Unconstrained Features: Neural Collapse for Shallow Neural Networks with General Datacited by 10
  14. 2025Neural Collapse is Globally Optimal in Deep Regularized ResNets and Transformerscited by 10
  15. 2023On the Robustness of Neural Collapse and the Neural Collapse of Robustnesscited by 9
  16. 2025Space Alignment Matters: The Missing Piece for Inducing Neural Collapse in Long-Tailed Learningcited by 2
  17. 2026Labeled TrustSet Guided: Batch Active Learning with Reinforcement Learningcited by 2
  18. 2026Active Learning Using Aggregated Acquisition Functions: Accuracy and Sustainability Analysiscited by 1
  19. 2026Neural collapse in the orthoplex regimecited by 1

Built on by

Not cited yet.