Research map / Probabilistic and scientific modeling / Variational and generative sampling
Mutual information estimation
Benchmarks and designs neural estimators of mutual information.
27 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2017–2026) · an approach in Variational and generative sampling, keyword cluster “estimation · probabilistic · variables”
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.
8 papers in 2023–24 and 5 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic shrank, from 1.9% to 1.0%.
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
- 2017 · Estimating Mutual Information for Discrete-Continuous Mixtures (NeurIPS 2017 · 106 citations)
- 2018 · Mutual Information Neural Estimation (ICML 2018 · 574 citations)
- 2020 · Understanding the Limitations of Variational Mutual Information Estimators (ICLR 2020 · 29 citations)
- 2021 · Estimating the Unique Information of Continuous Variables (NeurIPS 2021 · 10 citations)
- 2022 · Improving Mutual Information Estimation with Annealed and Energy-Based Bounds (ICLR 2022 · 1 citation)
- 2023 · Beyond Normal: On the Evaluation of Mutual Information Estimators (NeurIPS 2023 · 15 citations)
- 2024 · Differentiable Information Bottleneck for Deterministic Multi-view Clustering (CVPR 2024 · 27 citations)
- 2025 · Neural Mutual Information Estimation with Vector Copulas (NeurIPS 2025 · 0 citations)
- 2026 · Curse of Slicing: Why Sliced Mutual Information is a Deceptive Measure of Statistical Dependence (ICLR 2026 · 0 citations)
Earliest papers here
- Estimating Mutual Information for Discrete-Continuous Mixtures (NeurIPS 2017 · 106 citations)
- Nonparanormal Information Estimation (ICML 2017 · 8 citations)
- Mutual Information Neural Estimation (ICML 2018 · 574 citations)
What has been tried
23 papers here report 204 measured changes (a component added, removed or swapped, compared with a baseline): helped 67%, no clear effect 7%, hurt 16%, mixed 10%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| Add feature consistency (Lfea) to cluster consistency (Lclu) | MNIST-USPS, BDGP, and ESP · ACC | 85.72 -> 92.96; 82.58 -> 91.00; 48.01 -> 49.95 | Helped | Differentiable Information Bottleneck for Deterministic Multi-view Clustering (CVPR 2024 · 27 citations) |
| Use shift-based rather than random-based derangement | 5-step MI staircase, d=20, N=30k · time | 15 minutes -> less than 9 minutes | Helped | Mutual Information Estimation via $f$-Divergence and Data Derangements (NeurIPS 2024 · 4 citations) |
| Use a deeper critic rather than depth 1 | Gaussian, true MI 2 bits · MSE | 0.125 (depth 1); 0.117 (depth 2); 0.114 (depth 3); 0.114 (depth 4); 0.117 (depth 5) | Mixed | A Benchmark Suite for Evaluating Neural Mutual Information Estimators on Unstructured Datasets (NeurIPS 2024 · 0 citations) |
What reviewers will expect
Datasets and baselines named in the 6 papers of this approach from 2024–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Datasets
- MNIST · 2 of 6 papers
Baselines
- MINE · 4 of 6 papers
- InfoNCE · 3 of 6 papers
- NWJ · 3 of 6 papers
- D-V · 2 of 6 papers
- SMILE · 2 of 6 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 13 papers; from OpenAlex and the papers' affiliations).
- University of Cambridge · 2 papers
Use it
Download BibTeX of 26 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
- Likelihood-free and graphical model estimation (161 papers)
- Tractable probabilistic models (90 papers)
- Exact inference for probabilistic programs (103 papers)
- Latent variable marginalization and estimators (91 papers)
- Gradient estimation for discrete variables (74 papers)
Elsewhere in the topic
- Flow matching and flow sampling (144 papers)
- Posterior sampling with diffusion guidance (137 papers)
- Score-based generative modeling analysis (128 papers)
- Structured latent VAEs (125 papers)
- Disentangled and discrete latent models (92 papers)
- Variational inference theory and gradients (89 papers)
- Hamiltonian Monte Carlo (89 papers)
- Regularizing VAE latent information (89 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).