Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Normalization and unsupervised meta-learning
Adapts batch normalization for tasks and builds unsupervised or hard-sample meta-learning.
39 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2016–2026) · an approach in Zero-/few-shot and meta-learning, keyword cluster “maml · meta algorithms · meta meta”
State History Tried Preprints Who Export
Where it stands
Not enough benchmark data Fewer than four papers report the same benchmark, so there is no record to follow.
14 papers in 2023–24 and 4 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic shrank, from 6.8% to 2.3%.
How this is decided
A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.
How it got here
Most cited paper of each year
- 2016 · Learning from Small Sample Sets by Combining Unsupervised Meta-Training with CNNs (NeurIPS 2016 · 74 citations)
- 2017 · Learning to Model the Tail (NeurIPS 2017 · 259 citations)
- 2019 · Learning to Learn with Conditional Class Dependencies (ICLR 2019 · 55 citations)
- 2020 · MESA: Boost Ensemble Imbalanced Learning with MEta-SAmpler (NeurIPS 2020 · 31 citations)
- 2021 · MetaHTR: Towards Writer-Adaptive Handwritten Text Recognition (CVPR 2021 · 41 citations)
- 2022 · Meta Discovery: Learning to Discover Novel Classes given Very Limited Data (ICLR 2022 · 16 citations)
- 2023 · Diversity-Aware Meta Visual Prompting (CVPR 2023 · 60 citations)
- 2024 · FREE: Faster and Better Data-Free Meta-Learning (CVPR 2024 · 5 citations)
- 2025 · DataRater: Meta-Learned Dataset Curation (NeurIPS 2025 · 1 citation)
- 2026 · LiteGuard: Efficient Task-Agnostic Model Fingerprinting with Enhanced Generalization (ICLR 2026 · 0 citations)
Earliest papers here
- Learning from Small Sample Sets by Combining Unsupervised Meta-Training with CNNs (NeurIPS 2016 · 74 citations)
- Learning to Model the Tail (NeurIPS 2017 · 259 citations)
- Learning to Learn with Conditional Class Dependencies (ICLR 2019 · 55 citations)
What has been tried
32 papers here report 338 measured changes (a component added, removed or swapped, compared with a baseline): helped 68%, no clear effect 5%, hurt 23%, mixed 4%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| Add FIVE and BELL to the cross-task-replay baseline | CIFAR-FS 5-way 1-shot · accuracy | 36.33 ± 0.81 -> 39.13 ± 0.85 | Helped | FREE: Faster and Better Data-Free Meta-Learning (CVPR 2024 · 5 citations) |
| Remove the performance gradient update to the prediction model (P2P) | Crop classification, ablation study · LF | 0.112 -> 0.176 | Hurt | Referee-Meta-Learning for Fast Adaptation of Locational Fairness (AAAI 2024 · 4 citations) |
| use hard rather than easy validation samples in LRW | ImageNet-1K · accuracy | 83.17 -> 85.09 | Helped | Improving Generalization via Meta-Learning on Hard Samples (CVPR 2024 · 4 citations) |
| Include meta-learning of the performance and cost predictors | Meta-Album Micro, Mini, and Extended · normalized regret | Helped | Quick-Tune: Quickly Learning Which Pretrained Model to Finetune and How (ICLR 2024 · 3 citations) | |
| Use eight 400M inner models instead of 150M inner models | C4/noclean · relative performance improvement | Helped | DataRater: Meta-Learned Dataset Curation (NeurIPS 2025 · 1 citation) | |
| Add meta-weight-nets | GQA-CCG · accuracy | 62.07 -> 63.12 | Helped | Consistency of Compositional Generalization Across Multiple Levels (AAAI 2025 · 0 citations) |
| add adversarial perturbation on query images | Meta-Dataset 5-way 1-shot · average accuracy (%) | Helped | Mixture of Adversarial LoRAs: Boosting Robust Generalization in Meta-Tuning (NeurIPS 2024 · 0 citations) |
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 18 papers; from OpenAlex and the papers' affiliations).
- Zhejiang University · 2 papers
- University of Science and Technology of China · 2 papers
- Google · 2 papers
- Nanyang Technological University · 2 papers
Use it
Download BibTeX of 29 key papersOpen the topic in the interactive map
Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.
Related approaches
Same keyword cluster
- Meta-augmentation and meta-objectives (60 papers)
- Meta-learning analysis and benchmarks (50 papers)
Elsewhere in the topic
- Generative feature synthesis for zero-shot (117 papers)
- Zero-shot robustness and transfer (101 papers)
- Robust fine-tuning of zero-shot models (92 papers)
- Transductive and semantic-dictionary zero-shot (90 papers)
- Embedding adaptation for few-shot classification (71 papers)
- Fine-tuning strategies for few-shot (67 papers)
- Fine-grained and cross-domain few-shot baselines (59 papers)
- Alignment-based few-shot metric learning (38 papers)
Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).