Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Sparse variational Gaussian processes
Scales Gaussian processes with sparse inducing-point and variational approximations.
92 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2016–2026) · an approach in Uncertainty and Gaussian processes, keyword cluster “gaussian processes · gp · kernel”
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 14 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic shrank, from 6.6% to 4.5%.
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 · Stochastic Variational Deep Kernel Learning (NeurIPS 2016 · 97 citations)
- 2017 · Convolutional Gaussian Processes (NeurIPS 2017 · 51 citations)
- 2018 · GPyTorch: Blackbox Matrix-Matrix Gaussian Process Inference with GPU Acceleration (NeurIPS 2018 · 528 citations)
- 2019 · Rates of Convergence for Sparse Variational Gaussian Process Regression (ICML 2019 · 48 citations)
- 2020 · Scalable Variational Bayesian Kernel Selection for Sparse Gaussian Process Regression (AAAI 2020 · 19 citations)
- 2021 · Global inducing point variational posteriors for Bayesian neural networks and deep Gaussian processes (ICML 2021 · 17 citations)
- 2022 · Revisiting Active Sets for Gaussian Process Decoders (NeurIPS 2022 · 3 citations)
- 2023 · GAUCHE: A Library for Gaussian Processes in Chemistry (NeurIPS 2023 · 33 citations)
- 2024 · Sparse Variational Student-t Processes (AAAI 2024 · 5 citations)
- 2025 · New Bounds for Sparse Variational Gaussian Processes (ICML 2025 · 0 citations)
- 2026 · Sublinear Variational Optimization of Gaussian Mixture Models with Millions to Billions of Parameters (NeurIPS 2026 · 1 citation)
Earliest papers here
- Stochastic Variational Deep Kernel Learning (NeurIPS 2016 · 97 citations)
- Structured and Efficient Variational Deep Learning with Matrix Gaussian Posteriors (ICML 2016 · 88 citations)
- A Distributed Variational Inference Framework for Unifying Parallel Sparse Gaussian Process Regression Models (ICML 2016 · 33 citations)
What has been tried
80 papers here report 722 measured changes (a component added, removed or swapped, compared with a baseline): helped 67%, no clear effect 9%, hurt 18%, mixed 5%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| Switch the optimizer from SGD to Adam or LBFGS | Experiments · Performance | No clear effect | Domain Invariant Learning for Gaussian Processes and Bayesian Exploration (AAAI 2024 · 2 citations) | |
| Increase spatial resolution coarseness | Synthetic squared-exponential GP data · Wasserstein distance | Hurt | Numerically Stable Sparse Gaussian Processes via Minimum Separation using Cover Trees (NeurIPS 2024 · 1 citation) | |
| Optimize the additional scalar variance parameter v instead of fixing v=1 | NYBikes Poisson · test log-likelihood | M=8: -36.397(6.017) -> -9.713(0.345); M=16: -16.557(4.307) -> -9.301(0.296); M=32: -8.556(0.728) -> -8.203(0.190) | Helped | New Bounds for Sparse Variational Gaussian Processes (ICML 2025 · 0 citations) |
| Increase random feature count from 2M=20 to 2M=500 | 2D synthetic λ₂D¹.⁰ · L² | 147 -> 53.0 | Helped | K$^2$IE: Kernel Method-based Kernel Intensity Estimators for Inhomogeneous Poisson Processes (ICML 2025 · 0 citations) |
| Use oracle ordering rather than kernel-similarity ordering for the HiPPO recurrence | Skillcraft and Powerplant UCI datasets · NLPD | Helped | Recurrent Memory for Online Interdomain Gaussian Processes (NeurIPS 2025 · 0 citations) | |
| Increase the block size in block-diagonal variational scaling | kin40k, M=256 · RMSE | 0.217 -> 0.200 | Helped | Sparse Gaussian Processes: Structured Approximations and Power-EP Revisited (NeurIPS 2025 · 0 citations) |
| use semi-amortization | Rotated MNIST · MSE | 0.009 ± 0.0001 -> 0.010 ± 0.0004 | Hurt | Bayesian Basis Function Approximation for Scalable Gaussian Process Priors in Deep Generative Models (ICML 2025) |
| Increase inducing points from 2 to 20 | Cora · accuracy | Hurt | Amortized Variational Deep Kernel Learning (ICML 2024) | |
| use SubsCoRe-Bound instead of SubsCoRe-Center | Ising Hamiltonian, L=3, Q=5 · measurement-shot efficiency | Hurt | Adaptive Observation Cost Control for Variational Quantum Eigensolvers (ICML 2024) | |
| Increase DGP depth from L=4 to L=5 | SUSY test · AUC | 0.887 -> 0.886 | Hurt | Sparse Inducing Points in Deep Gaussian Processes: Enhancing Modeling with Denoising Diffusion Variational Inference (ICML 2024) |
| Reduce mantissa precision from 14 bits to 8 bits | California Housing (ch) · ΔTest RMSE | −0.02 to 0.14 | Hurt | Energy-Efficient Gaussian Processes Using Low-Precision Arithmetic (ICML 2024) |
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.
Baselines
- Exact GP · 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 32 papers; from OpenAlex and the papers' affiliations).
- University of Cambridge · 3 papers
- South China University of Technology · 2 papers
- Imperial College London · 2 papers
- Aalto University · 2 papers
- Google DeepMind · 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
- PAC-Bayes and decoupled GP learning (83 papers)
- Infinite-width networks as Gaussian processes (49 papers)
- Structured Gaussian process models (39 papers)
Elsewhere in the topic
- Nonparametric and PAC-Bayes Bayesian learning (130 papers)
- Simulation-based inference (118 papers)
- Priors and functional variational uncertainty (112 papers)
- Variational Bayesian deep learning (105 papers)
- Approximate inference in Bayesian neural networks (93 papers)
- Epistemic versus aleatoric decomposition (75 papers)
- Single-model predictive uncertainty (74 papers)
- Bayesian view of meta-learning (63 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).