mlsanityData to 7 Oct 2026

k-NN as a Simple and Effective Estimator of Transferability

arXiv 2503.18528 ↗First posted 24 Mar 2025cs.LG cs.CVcited by 1

Not on the map. Why?

Abstract (arXiv)

How well can one expect transfer learning to work in a new setting where the domain is shifted, the task is different, and the architecture changes? Many transfer learning metrics have been proposed to answer this question. But how accurate are their predictions in a realistic new setting? We conducted an extensive evaluation involving over 42,000 experiments comparing 23 transferability metrics across 16 different datasets to assess their ability to predict transfer performance. Our findings reveal that none of the existing metrics perform well across the board. However, we find that a simple k-nearest neighbor evaluation -- as is commonly used to evaluate feature quality for self-supervision -- not only surpasses existing metrics, but also offers better computational efficiency and ease of implementation.

Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 56 of its 88 references that are in the map. First posted on arXiv 24 Mar 2025.

Builds on

56 of 88 references.

References · 56

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  3. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,082
  4. 2014ImageNet Large Scale Visual Recognition Challengecited by 8,441
  5. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,482
  6. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,893
  7. 2016Densely Connected Convolutional Networkscited by 6,691
  8. 2015Rethinking the Inception Architecture for Computer Visioncited by 5,451
  9. 2018MobileNetV2: Inverted Residuals and Linear Bottleneckscited by 4,773
  10. 2020Training data-efficient image transformers & distillation through attentioncited by 4,560
  11. 2015Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classificationcited by 3,590
  12. 2016Wide Residual Networkscited by 3,057
38 more
  1. 2016Aggregated Residual Transformations for Deep Neural Networkscited by 2,540
  2. 2013Describing Textures in the Wildcited by 2,110
  3. 2021Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutionscited by 1,960
  4. 2018Unsupervised Feature Learning via Non-Parametric Instance-level Discriminationcited by 1,838
  5. 2013Fine-Grained Visual Classification of Aircraftcited by 1,701
  6. 2018Moment Matching for Multi-Source Domain Adaptationcited by 1,327
  7. 2018Demystifying MMD GANscited by 1,268
  8. 2014How transferable are features in deep neural networks?cited by 1,216
  9. 2020Designing Network Design Spacescited by 954
  10. 2018Taskonomy: Disentangling Task Transfer Learningcited by 631
  11. 2018Do Better ImageNet Models Transfer Better?cited by 560
  12. 2018Rethinking ImageNet Pre-trainingcited by 376
  13. 2019Transfusion: Understanding Transfer Learning for Medical Imagingcited by 366
  14. 2020What is being transferred in transfer learning?cited by 365
  15. 2021Rethinking Spatial Dimensions of Vision Transformerscited by 361
  16. 2016AID: A Benchmark Dataset for Performance Evaluation of Aerial Scene Classificationcited by 343
  17. 2019Task2Vec: Task Embedding for Meta-Learningcited by 197
  18. 2020LEEP: A New Measure to Evaluate Transferability of Learned Representationscited by 182
  19. 2018Large Scale Fine-Grained Categorization and Domain-Specific Transfer Learningcited by 174
  20. 2021LogME: Practical Assessment of Pre-trained Models for Transfer Learningcited by 168
  21. 2020Geometric Dataset Distances via Optimal Transportcited by 157
  22. 2019Transferability and Hardness of Supervised Classification Taskscited by 125
  23. 2020A Survey on Negative Transfercited by 120
  24. 2022An Information-Theoretic Approach to Transferability in Task Transfer Learningcited by 79
  25. 2021OTCE: A Transferability Metric for Cross-Domain Cross-Task Representationscited by 62
  26. 2022What Makes Transfer Learning Work For Medical Images: Feature Reuse & Other Factorscited by 61
  27. 2021Factors of Influence for Transfer Learning across Diverse Appearance Domains and Task Typescited by 54
  28. 2021Transferability Estimation using Bhattacharyya Class Separabilitycited by 53
  29. 2020Ranking Neural Checkpointscited by 50
  30. 2021Frustratingly Easy Transferability Estimationcited by 44
  31. 2014A Weighted Correlation Index for Rankings with Tiescited by 41
  32. 2021A linearized framework and a new benchmark for model selection for fine-tuningcited by 41
  33. 2021Scalable Diverse Model Selection for Accessible Transfer Learningcited by 40
  34. 2019The Shape of Data: Intrinsic Distance for Data Distributionscited by 38
  35. 2022PACTran: PAC-Bayesian Metrics for Estimating the Transferability of Pretrained Models to Classification Taskscited by 30
  36. 2022How stable are Transferability Metrics evaluations?cited by 30
  37. 2022Not All Models Are Equal: Predicting Model Transferability in a Self-challenging Fisher Spacecited by 30
  38. 2019Deep Model Transferability from Attribution Mapscited by 28

50 most cited of 56 All 56 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026A Source Domain is All You Need: Source-Only Cross-OS Transfer Learning for APT Anomaly Detection via Semantic Alignment and Optimal Transportcited by 0