mlsanityData to 7 Oct 2026

LEEP: A New Measure to Evaluate Transferability of Learned Representations

arXiv 2002.12462 ↗First posted 27 Feb 2020cs.LG cs.CV stat.MLcited by 182

Not on the map. Why?

Abstract (arXiv)

We introduce a new measure to evaluate the transferability of representations learned by classifiers. Our measure, the Log Expected Empirical Prediction (LEEP), is simple and easy to compute: when given a classifier trained on a source data set, it only requires running the target data set through this classifier once. We analyze the properties of LEEP theoretically and demonstrate its effectiveness empirically. Our analysis shows that LEEP can predict the performance and convergence speed of both transfer and meta-transfer learning methods, even for small or imbalanced data. Moreover, LEEP outperforms recently proposed transferability measures such as negative conditional entropy and H scores. Notably, when transferring from ImageNet to CIFAR100, LEEP can achieve up to 30% improvement compared to the best competing method in terms of the correlations with actual transfer accuracy.

Key facts. Cited by 182 papers in the map (data through 7 Oct 2026). First posted on arXiv 27 Feb 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 22; 2022: 29; 2023: 33; 2024: 31; 2025: 34; 2026: 33.

182 papers.

Sort:

By area:

  1. 2021LogME: Practical Assessment of Pre-trained Models for Transfer Learningcited by 168
  2. 2022Deep Model Reassemblycited by 78
  3. 2023Sensitivity-Aware Visual Parameter-Efficient Fine-Tuningcited by 74
  4. 2021OTCE: A Transferability Metric for Cross-Domain Cross-Task Representationscited by 62
  5. 2021What to Pre-Train on? Efficient Intermediate Task Selectioncited by 62
  6. 2022Transferability in Deep Learning: A Surveycited by 57
  7. 2021Transferability Estimation using Bhattacharyya Class Separabilitycited by 53
  8. 2024Scaling Laws for Downstream Task Performance of Large Language Modelscited by 49
  9. 2021Frustratingly Easy Transferability Estimationcited by 44
  10. 2022Memory Efficient Continual Learning with Transformerscited by 43
  11. 2021A linearized framework and a new benchmark for model selection for fine-tuningcited by 41
  12. 2021Scalable Diverse Model Selection for Accessible Transfer Learningcited by 40
38 more
  1. 2021Zoo-Tuning: Adaptive Transfer from a Zoo of Modelscited by 35
  2. 2024Selecting Large Language Model to Fine-tune via Rectified Scaling Lawcited by 33
  3. 2023Model Spider: Learning to Rank Pre-Trained Models Efficientlycited by 32
  4. 2022PACTran: PAC-Bayesian Metrics for Estimating the Transferability of Pretrained Models to Classification Taskscited by 30
  5. 2022How stable are Transferability Metrics evaluations?cited by 30
  6. 2022Not All Models Are Equal: Predicting Model Transferability in a Self-challenging Fisher Spacecited by 30
  7. 2021Ranking and Tuning Pre-trained Models: A New Paradigm for Exploiting Model Hubscited by 28
  8. 2023Scalable Optimal Transport Methods in Machine Learning: A Contemporary Surveycited by 27
  9. 2025Capability Instruction Tuning: A New Paradigm for Dynamic LLM Routingcited by 25
  10. 2025We Should Chart an Atlas of All the World's Modelscited by 25
  11. 2021Transferability Metrics for Selecting Source Model Ensemblescited by 24
  12. 2022Wasserstein Task Embedding for Measuring Task Similaritiescited by 24
  13. 2024Wings: Learning Multimodal LLMs without Text-only Forgettingcited by 23
  14. 2023Identification of Negative Transfers in Multitask Learning Using Surrogate Modelscited by 20
  15. 2025How Well Do Supervised 3D Models Transfer to Medical Imaging Tasks?cited by 19
  16. 2021A Multiplexed Network for End-to-End, Multilingual OCRcited by 18
  17. 2023ETran: Energy-Based Transferability Estimationcited by 18
  18. 2022Continual Learning with Transformers for Image Classificationcited by 17
  19. 2023The Role of Pre-training Data in Transfer Learningcited by 17
  20. 2023Quick-Tune: Quickly Learning Which Pretrained Model to Finetune and Howcited by 17
  21. 2023Foundation Model is Efficient Multimodal Multitask Model Selectorcited by 16
  22. 2021Newer is not always better: Rethinking transferability metrics, their peculiarities, stability and performancecited by 15
  23. 2023Can We Evaluate Domain Adaptation Models Without Target-Domain Labels?cited by 14
  24. 2022Deep transfer learning for image classification: a surveycited by 13
  25. 2022Transferability-Guided Cross-Domain Cross-Task Transfer Learningcited by 13
  26. 2022Uncertainty-Induced Transferability Representation for Source-Free Unsupervised Domain Adaptationcited by 13
  27. 2024Which Model to Transfer? A Survey on Transferability Estimationcited by 13
  28. 2022ZooD: Exploiting Model Zoo for Out-of-Distribution Generalizationcited by 12
  29. 2024Bayesian-guided Label Mapping for Visual Reprogrammingcited by 12
  30. 2023Understanding the Transferability of Representations via Task-Relatednesscited by 11
  31. 2022Beyond Static Models and Test Sets: Benchmarking the Potential of Pre-trained Models Across Tasks and Languagescited by 10
  32. 2023How to Estimate Model Transferability of Pre-Trained Speech Models?cited by 10
  33. 2023Pick the Best Pre-trained Model: Towards Transferability Estimation for Medical Image Segmentationcited by 10
  34. 2023Exploring Model Transferability through the Lens of Potential Energycited by 10
  35. 2022Task Discovery: Finding the Tasks that Neural Networks Generalize oncited by 9
  36. 2023Fast and Accurate Transferability Measurement by Evaluating Intra-class Feature Variancecited by 9
  37. 2024Enhancing Trust in LLMs: Algorithms for Comparing and Interpreting LLMscited by 9
  38. 2021GAN Cocktail: mixing GANs without dataset accesscited by 8

50 most cited of 182 All 182 (JSON)