Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Bayesian view of meta-learning
Recasts meta-learning and in-context learning as hierarchical Bayesian inference.
63 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2016–2026) · an approach in Uncertainty and Gaussian processes, keyword cluster “bayesian inference · posterior · priors”
State History Tried Checklist 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.
18 papers in 2023–24 and 17 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic shrank, from 6.6% to 5.5%.
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 under uncertainty: a comparison between R-W and Bayesian approach (NeurIPS 2016 · 1 citation)
- 2017 · Deep Bayesian Active Learning with Image Data (ICML 2017 · 573 citations)
- 2018 · Bayesian Model-Agnostic Meta-Learning (NeurIPS 2018 · 156 citations)
- 2019 · Amortized Bayesian Meta-Learning (ICLR 2019 · 71 citations)
- 2020 · Variational Metric Scaling for Metric-Based Meta-Learning (AAAI 2020 · 40 citations)
- 2021 · PACOH: Bayes-Optimal Meta-Learning with PAC-Guarantees (ICML 2021 · 38 citations)
- 2022 · Tracking Functional Changes in Nonstationary Signals with Evolutionary Ensemble Bayesian Model for Robust Neural Decoding (NeurIPS 2022 · 2 citations)
- 2023 · BETA-CD: A Bayesian Meta-Learned Cognitive Diagnosis Framework for Personalized Learning (AAAI 2023 · 21 citations)
- 2024 · RECOMBINER: Robust and Enhanced Compression with Bayesian Implicit Neural Representations (ICLR 2024 · 2 citations)
- 2025 · Inverse decision-making using neural amortized Bayesian actors (ICLR 2025 · 1 citation)
- 2026 · BALLAST: Bayesian Active Learning with Look-ahead Amendment for Sea-drifter Trajectories under Spatio-Temporal Vector Fields (ICML 2026 · 0 citations)
Earliest papers here
- Learning under uncertainty: a comparison between R-W and Bayesian approach (NeurIPS 2016 · 1 citation)
- Deep Bayesian Active Learning with Image Data (ICML 2017 · 573 citations)
- Robust and Efficient Transfer Learning with Hidden Parameter Markov Decision Processes (NeurIPS 2017 · 54 citations)
What has been tried
43 papers here report 418 measured changes (a component added, removed or swapped, compared with a baseline): helped 64%, no clear effect 11%, hurt 19%, mixed 6%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| add linear weight reparameterization | CIFAR-10 · PSNR | up to 4dB gain | Helped | RECOMBINER: Robust and Enhanced Compression with Bayesian Implicit Neural Representations (ICLR 2024 · 2 citations) |
| Fix one of the confounding parameters at its true value instead of inferring both | Synthetic asymmetric quadratic cost · MSE of inferred parameters | Helped | Inverse decision-making using neural amortized Bayesian actors (ICLR 2025 · 1 citation) | |
| Replace the Gaussian posterior with a CNF posterior | Omniglot→EMNIST 1-shot · accuracy | 80.95 -> 72.02 | Hurt | Hypernetwork Approach to Bayesian MAML (Student Abstract) (AAAI 2025 · 0 citations) |
| Use flow matching rather than diffusion objectives | Evaluated scenarios · posterior sample similarity | Helped | Can Transformers Learn Full Bayesian Inference in Context? (ICML 2025 · 0 citations) | |
| Remove the assumed computational constraints from the functional form | Transformer next-token predictions · prediction accuracy | Hurt | In-Context Learning Strategies Emerge Rationally (NeurIPS 2025 · 0 citations) | |
| Remove positional encodings | Dense regression, length generalization · error | Helped | In-Context Learning through the Bayesian Prism (ICLR 2024 · 0 citations) | |
| replace Gaussian prior with Spike-and-Slab prior | Average (eight datasets), ViT-B/32 · accuracy | 84.32% -> 84.69% | Helped | Variational Task Vector Composition (NeurIPS 2025) |
What reviewers will expect
Datasets and baselines named in the 9 papers of this approach from 2025–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Datasets
- MNIST · 2 of 9 papers
- SVHN · 2 of 9 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 35 papers; from OpenAlex and the papers' affiliations).
- University of Science and Technology of China · 3 papers
- Google DeepMind · 3 papers
- Samsung · 2 papers
- University of Cambridge · 2 papers
- United Imaging Healthcare · 2 papers
- University of Amsterdam · 2 papers
- Stanford University · 2 papers
Use it
Download BibTeX of 40 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
- Nonparametric and PAC-Bayes Bayesian learning (130 papers)
- Simulation-based inference (118 papers)
- Variational Bayesian deep learning (105 papers)
- Approximate inference in Bayesian neural networks (93 papers)
Elsewhere in the topic
- Priors and functional variational uncertainty (112 papers)
- Sparse variational Gaussian processes (92 papers)
- PAC-Bayes and decoupled GP learning (83 papers)
- Epistemic versus aleatoric decomposition (75 papers)
- Single-model predictive uncertainty (74 papers)
- Bayesian decision-making under uncertainty (57 papers)
- Infinite-width networks as Gaussian processes (49 papers)
- Regression uncertainty quantification (43 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).