mlsanityData to 7 Oct 2026

MT3: Meta Test-Time Training for Self-Supervised Test-Time Adaption

arXiv 2103.16201 ↗First posted 30 Mar 2021cs.CV cs.AIcited by 65

Not on the map. Why?

Abstract (arXiv)

An unresolved problem in Deep Learning is the ability of neural networks to cope with domain shifts during test-time, imposed by commonly fixing network parameters after training. Our proposed method Meta Test-Time Training (MT3), however, breaks this paradigm and enables adaption at test-time. We combine meta-learning, self-supervision and test-time training to learn to adapt to unseen test distributions. By minimizing the self-supervised loss, we learn task-specific model parameters for different tasks. A meta-model is optimized such that its adaption to the different task-specific models leads to higher performance on those tasks. During test-time a single unlabeled image is sufficient to adapt the meta-model parameters. This is achieved by minimizing only the self-supervised loss component resulting in a better prediction for that image. Our approach significantly improves the state-of-the-art results on the CIFAR-10-Corrupted image classification benchmark. Our implementation is available on GitHub.

Key facts. Cited by 65 papers in the map (data through 7 Oct 2026). Builds on 35 of its 42 references that are in the map. First posted on arXiv 30 Mar 2021.

Builds on

35 of 42 references.

References · 35

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 11,979
  3. 2018Representation Learning with Contrastive Predictive Codingcited by 7,662
  4. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,482
  5. 2013Intriguing properties of neural networkscited by 5,114
  6. 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,861
  7. 2020Bootstrap your own latent: A new approach to self-supervised Learningcited by 4,712
  8. 2019Benchmarking Neural Network Robustness to Common Corruptions and Perturbationscited by 2,857
  9. 2020The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalizationcited by 1,842
  10. 2018Unsupervised Representation Learning by Predicting Image Rotationscited by 1,586
  11. 2018Group Normalizationcited by 1,514
  12. 2019Do ImageNet Classifiers Generalize to ImageNet?cited by 1,507
23 more
  1. 2020Big Self-Supervised Models are Strong Semi-Supervised Learnerscited by 1,375
  2. 2018On First-Order Meta-Learning Algorithmscited by 1,003
  3. 2020Meta-Learning in Neural Networks: A Surveycited by 944
  4. 2020Out-of-Distribution Generalization via Risk Extrapolation (REx)cited by 737
  5. 2018Deep Visual Domain Adaptation: A Surveycited by 634
  6. 2017Adversarial Examples Are Not Easily Detected: Bypassing Ten Detection Methodscited by 542
  7. 2018A Survey on Deep Transfer Learningcited by 417
  8. 2017Meta-SGD: Learning to Learn Quickly for Few-Shot Learningcited by 377
  9. 2014Deeply-Supervised Netscited by 364
  10. 2018A Survey of Unsupervised Deep Domain Adaptationcited by 361
  11. 2018How to train your MAMLcited by 306
  12. 2018Why do deep convolutional networks generalize so poorly to small image transformations?cited by 198
  13. 2020Adversarial Robustness: From Self-Supervised Pre-Training to Fine-Tuningcited by 152
  14. 2020A Review of Single-Source Deep Unsupervised Visual Domain Adaptationcited by 96
  15. 2020Deep Visual Domain Adaptationcited by 91
  16. 2018Unsupervised Learning via Meta-Learningcited by 81
  17. 2019When Does Self-supervision Improve Few-shot Learning?cited by 78
  18. 2018Unsupervised Meta-Learning For Few-Shot Image Classificationcited by 74
  19. 2020Adversarial Style Mining for One-Shot Unsupervised Domain Adaptationcited by 64
  20. 2018Meta-Learning Update Rules for Unsupervised Representation Learningcited by 37
  21. 2018One-Shot Unsupervised Cross Domain Translationcited by 35
  22. 2020Learn2Perturb: an End-to-end Feature Perturbation Learning to Improve Adversarial Robustnesscited by 27
  23. 2020Improving out-of-distribution generalization via multi-task self-supervised pretrainingcited by 26

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 4; 2022: 11; 2023: 16; 2024: 14; 2025: 18; 2026: 2.

65 papers.

Sort:

By area:

  1. 2022Efficient Test-Time Model Adaptation without Forgettingcited by 467
  2. 2023Towards Stable Test-Time Adaptation in Dynamic Wild Worldcited by 409
  3. 2021MEMO: Test Time Robustness via Adaptation and Augmentationcited by 376
  4. 2023A Comprehensive Survey on Test-Time Adaptation under Distribution Shiftscited by 353
  5. 2024Entropy is not Enough for Test-Time Adaptation: From the Perspective of Disentangled Factorscited by 106
  6. 2022Robust Mean Teacher for Continual and Gradual Test-Time Adaptationcited by 102
  7. 2024Test-Time Model Adaptation with Only Forward Passescited by 74
  8. 2023Universal Test-time Adaptation through Weight Ensembling, Diversity Weighting, and Prior Correctioncited by 65
  9. 2021SITA: Single Image Test-time Adaptationcited by 57
  10. 2025Test-Time Learning for Large Language Modelscited by 46
  11. 2023Improved Test-Time Adaptation for Domain Generalizationcited by 45
  12. 2025A Survey of Test-Time Compute: From Intuitive Inference to Deliberate Reasoningcited by 40
38 more
  1. 2021If your data distribution shifts, use self-learningcited by 36
  2. 2022TTTFlow: Unsupervised Test-Time Training with Normalizing Flowcited by 33
  3. 2023Diffusion-TTA: Test-time Adaptation of Discriminative Models via Generative Feedbackcited by 27
  4. 2024Protected Test-Time Adaptation via Online Entropy Matching: A Betting Approachcited by 22
  5. 2024Uncertainty-Calibrated Test-Time Model Adaptation without Forgettingcited by 20
  6. 2022Test-Time Training for Graph Neural Networkscited by 17
  7. 2023What, How, and When Should Object Detectors Update in Continually Changing Test Domains?cited by 15
  8. 2022Self-Adaptive Forecasting for Improved Deep Learning on Non-Stationary Time-Seriescited by 14
  9. 2023Improving Domain Generalization with Domain Relationscited by 14
  10. 2022Learning Instance-Specific Adaptation for Cross-Domain Segmentationcited by 13
  11. 2024Distribution Alignment for Fully Test-Time Adaptation with Dynamic Online Data Streamscited by 13
  12. 2025Continuous Self-Improvement of Large Language Models by Test-time Training with Verifier-Driven Sample Selectioncited by 13
  13. 2022Introducing Intermediate Domains for Effective Self-Training during Test-Timecited by 12
  14. 2023Uncovering Adversarial Risks of Test-Time Adaptationcited by 12
  15. 2023Towards Task Sampler Learning for Meta-Learningcited by 11
  16. 2024Towards Robust and Efficient Cloud-Edge Elastic Model Adaptation via Selective Entropy Distillationcited by 11
  17. 2025Beyond Entropy: Region Confidence Proxy for Wild Test-Time Adaptationcited by 11
  18. 2022Test-time Adaptation with Slot-Centric Modelscited by 9
  19. 2025Test-Time Model Adaptation for Quantized Neural Networkscited by 9
  20. 2022TTAPS: Test-Time Adaption by Aligning Prototypes using Self-Supervisioncited by 8
  21. 2023Robust Question Answering against Distribution Shifts with Test-Time Adaptation: An Empirical Studycited by 8
  22. 2023Bag of Tricks for Fully Test-Time Adaptationcited by 8
  23. 2021Re-using Adversarial Mask Discriminators for Test-time Training under Distribution Shiftscited by 7
  24. 2024FastMem: Fast Memorization of Prompt Improves Context Awareness of Large Language Modelscited by 7
  25. 2025Test-Time Visual In-Context Tuningcited by 7
  26. 2025Self-Bootstrapping for Versatile Test-Time Adaptationcited by 7
  27. 2022Understanding the Tricks of Deep Learning in Medical Image Segmentation: Challenges and Future Directionscited by 5
  28. 2025Progressive Test Time Energy Adaptation for Medical Image Segmentationcited by 5
  29. 2023Is it an i or an l: Test-time Adaptation of Text Line Recognition Modelscited by 3
  30. 2023Single Image Test-Time Adaptation for Segmentationcited by 3
  31. 2023Test-Time Training for Semantic Segmentation with Output Contrastive Losscited by 3
  32. 2025DocTTT: Test-Time Training for Handwritten Document Recognition Using Meta-Auxiliary Learningcited by 3
  33. 2025Test-time Correlation Alignmentcited by 3
  34. 2025Adapt in the Wild: Test-Time Entropy Minimization with Sharpness and Feature Regularizationcited by 3
  35. 2025ZeroSiam: An Efficient Asymmetry for Test-Time Entropy Optimization without Collapsecited by 3
  36. 2023Enhancing Plasticity for First Session Adaptation Continual Learningcited by 2
  37. 2024Test-time Adaptation Meets Image Enhancement: Improving Accuracy via Uncertainty-aware Logit Switchingcited by 2
  38. 2025Exploring Structured Semantic Priors Underlying Diffusion Score for Test-time Adaptationcited by 2

50 most cited of 65 All 65 (JSON)