mlsanityData to 7 Oct 2026

Meta-Learning with Latent Embedding Optimization

arXiv 1807.05960 ↗First posted 16 Jul 2018cs.LG cs.CV stat.MLcited by 476

Not on the map. Why?

Abstract (arXiv)

Gradient-based meta-learning techniques are both widely applicable and proficient at solving challenging few-shot learning and fast adaptation problems. However, they have practical difficulties when operating on high-dimensional parameter spaces in extreme low-data regimes. We show that it is possible to bypass these limitations by learning a data-dependent latent generative representation of model parameters, and performing gradient-based meta-learning in this low-dimensional latent space. The resulting approach, latent embedding optimization (LEO), decouples the gradient-based adaptation procedure from the underlying high-dimensional space of model parameters. Our evaluation shows that LEO can achieve state-of-the-art performance on the competitive miniImageNet and tieredImageNet few-shot classification tasks. Further analysis indicates LEO is able to capture uncertainty in the data, and can perform adaptation more effectively by optimizing in latent space.

Key facts. Cited by 476 papers in the map (data through 7 Oct 2026). First posted on arXiv 16 Jul 2018.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2020: 1; 2021: 162; 2022: 127; 2023: 89; 2024: 53; 2025: 28; 2026: 16.

476 papers.

Sort:

By area:

  1. 2022Transformers learn in-context by gradient descentcited by 714
  2. 2021FedBABU: Towards Enhanced Representation for Federated Image Classificationcited by 180
  3. 2021FSCE: Few-Shot Object Detection via Contrastive Proposal Encodingcited by 168
  4. 2021Free Lunch for Few-shot Learning: Distribution Calibrationcited by 165
  5. 2022A Comprehensive Survey of Few-shot Learning: Evolution, Applications, Challenges, and Opportunitiescited by 162
  6. 2021Investigating Bi-Level Optimization for Learning and Vision from a Unified Perspective: A Survey and Beyondcited by 151
  7. 2021Using Shape to Categorize: Low-Shot Learning with an Explicit Shape Biascited by 115
  8. 2023Open-domain Visual Entity Recognition: Towards Recognizing Millions of Wikipedia Entitiescited by 112
  9. 2023A Brief Review of Hypernetworks in Deep Learningcited by 102
  10. 2021On the Role of Neural Collapse in Transfer Learningcited by 100
  11. 2022Self-Support Few-Shot Semantic Segmentationcited by 96
  12. 2021Self-Promoted Prototype Refinement for Few-Shot Class-Incremental Learningcited by 95
38 more
  1. 2020Offline Meta-Reinforcement Learning with Advantage Weightingcited by 92
  2. 2021Generalized Few-Shot Object Detection without Forgettingcited by 88
  3. 2022Learning from Few Examples: A Summary of Approaches to Few-Shot Learningcited by 86
  4. 2023FeCAM: Exploiting the Heterogeneity of Class Distributions in Exemplar-Free Continual Learningcited by 86
  5. 2022Tutorial on amortized optimizationcited by 83
  6. 2021Cross-domain Few-shot Learning with Task-specific Adapterscited by 79
  7. 2021Relational Embedding for Few-Shot Classificationcited by 74
  8. 2021FLEX: Unifying Evaluation for Few-Shot NLPcited by 67
  9. 2023Advances and Challenges in Meta-Learning: A Technical Reviewcited by 65
  10. 2022Transferability in Deep Learning: A Surveycited by 57
  11. 2021Exploring Complementary Strengths of Invariant and Equivariant Representations for Few-Shot Learningcited by 56
  12. 2021Learning Meta-class Memory for Few-Shot Semantic Segmentationcited by 56
  13. 2022HyperTransformer: Model Generation for Supervised and Semi-Supervised Few-Shot Learningcited by 55
  14. 2021Rectifying the Shortcut Learning of Background for Few-Shot Learningcited by 51
  15. 2021Boosting the Generalization Capability in Cross-Domain Few-shot Learning via Noise-enhanced Supervised Autoencodercited by 51
  16. 2021StyleMeUp: Towards Style-Agnostic Sketch-Based Image Retrievalcited by 50
  17. 2022Singular Value Fine-tuning: Few-shot Segmentation requires Few-parameters Fine-tuningcited by 50
  18. 2021Meta-Learning with Task-Adaptive Loss Function for Few-Shot Learningcited by 49
  19. 2022Generalizing to New Physical Systems via Context-Informed Dynamics Modelcited by 49
  20. 2021Shallow Bayesian Meta Learning for Real-World Few-Shot Recognitioncited by 48
  21. 2021Meta-FDMixup: Cross-Domain Few-Shot Learning Guided by Labeled Target Datacited by 48
  22. 2023A Survey on Few-Shot Class-Incremental Learningcited by 48
  23. 2023A Closer Look at Few-shot Classification Againcited by 46
  24. 2021Meta-Learning with Fewer Tasks through Task Interpolationcited by 45
  25. 2021Bootstrapped Meta-Learningcited by 45
  26. 2022Margin-Based Few-Shot Class-Incremental Learning with Class-Level Overfitting Mitigationcited by 45
  27. 2021Learning Dynamic Alignment via Meta-filter for Few-shot Learningcited by 41
  28. 2021Property-Aware Relation Networks for Few-Shot Molecular Property Predictioncited by 41
  29. 2023CLIP-guided Prototype Modulating for Few-shot Action Recognitioncited by 41
  30. 2021One-Shot Object Affordance Detection in the Wildcited by 40
  31. 2023MetaDiff: Meta-Learning with Conditional Diffusion for Few-Shot Learningcited by 40
  32. 2022Hypernetworks in Meta-Reinforcement Learningcited by 39
  33. 2024Meta-Unlearning on Diffusion Models: Preventing Relearning Unlearned Conceptscited by 39
  34. 2024Diffusion-Based Neural Network Weights Generationcited by 38
  35. 2021Meta-Learning Dynamics Forecasting Using Task Inferencecited by 37
  36. 2022Few-Shot Diffusion Modelscited by 37
  37. 2021Partner-Assisted Learning for Few-Shot Image Classificationcited by 36
  38. 2022Rethinking Generalization in Few-Shot Classificationcited by 36

50 most cited of 476 All 476 (JSON)