Research map / Probabilistic and scientific modeling / Molecules and proteins
Probabilistic embeddings and relational inference
Uses box embeddings, analogical embeddings and lifted inference for relational data.
93 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2016–2026) · an approach in Molecules and proteins, keyword cluster “latent · space · representations”
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.
21 papers in 2023–24 and 31 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic shrank, from 4.3% to 3.3%.
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 · Complex Embeddings for Simple Link Prediction (ICML 2016 · 1,122 citations)
- 2017 · Analogical Inference for Multi-relational Embeddings (ICML 2017 · 240 citations)
- 2018 · Classical Planning in Deep Latent Space: Bridging the Subsymbolic-Symbolic Boundary (AAAI 2018 · 110 citations)
- 2019 · Smoothing the Geometry of Probabilistic Box Embeddings (ICLR 2019 · 46 citations)
- 2020 · A Theory of Usable Information under Computational Constraints (ICLR 2020 · 30 citations)
- 2021 · Nested Named Entity Recognition with Partially-Observed TreeCRFs (AAAI 2021 · 53 citations)
- 2022 · A Probabilistic Graphical Model Based on Neural-Symbolic Reasoning for Visual Relationship Detection (CVPR 2022 · 19 citations)
- 2023 · Conditional Syntax Splitting for Non-monotonic Inference Operators (AAAI 2023 · 20 citations)
- 2024 · On the Affinity, Rationality, and Diversity of Hierarchical Topic Modeling (AAAI 2024 · 9 citations)
- 2025 · SPARTAN: A Sparse Transformer World Model Attending to What Matters (NeurIPS 2025 · 1 citation)
- 2026 · I Predict Therefore I Am: Is Next Token Prediction Enough to Learn Human-Interpretable Concepts from Data? (ICLR 2026 · 1 citation)
Earliest papers here
- Complex Embeddings for Simple Link Prediction (ICML 2016 · 1,122 citations)
- Hashtag-Based Sub-Event Discovery Using Mutually Generative LDA in Twitter (AAAI 2016 · 57 citations)
- Probing the Compositionality of Intuitive Functions (NeurIPS 2016 · 23 citations)
What has been tried
64 papers here report 595 measured changes (a component added, removed or swapped, compared with a baseline): helped 71%, no clear effect 10%, hurt 14%, 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 |
|---|---|---|---|---|
| Add TPD to TraCo | Wikitext-103 · PCC | -0.286 -> 0.167 | Helped | On the Affinity, Rationality, and Diversity of Hierarchical Topic Modeling (AAAI 2024 · 9 citations) |
| Add open-space dispersion and structuring | Skin cancer lesions · inter-class angular separation | Helped | Seeing Unseen: Discover Novel Biomedical Concepts via Geometry-Constrained Probabilistic Modeling (CVPR 2024 · 7 citations) | |
| Add MLPs | FN1.7 without exemplar · F1-score | 74.76 -> 74.84 | Helped | Frame Semantic Role Labeling Using Arbitrary-Order Conditional Random Fields (AAAI 2024 · 3 citations) |
| increase optimization steps from 10 to 40 | Matrix Inverse, harder difficulty · MSE | 0.2110 -> 0.2063 | Helped | Learning Iterative Reasoning through Energy Diffusion (ICML 2024 · 3 citations) |
| add discrete diffusion to PQ semantic IDs | Scientific · Recall@10 | 0.1126 -> 0.1207 | Helped | Breaking Determinism: Fuzzy Modeling of Sequential Recommendation Using Discrete State Space Diffusion Model (NeurIPS 2024 · 2 citations) |
| Add an invertible neural network | CelebA attractive, male · EO | 33.4 -> 26.5 | Helped | Constructing Fair Latent Space for Intersection of Fairness and Explainability (AAAI 2025 · 0 citations) |
| add autoregressive skill prediction to CST | LIBERO · overall success rate | 89.5% vs 93.6% | Helped | STAR: Learning Diverse Robot Skill Abstractions through Rotation-Augmented Vector Quantization (ICML 2025 · 0 citations) |
| use sequential, history-aware slot prediction instead of simultaneous slot prediction | NBA · ADE/FDE | 0.62/0.94 -> 0.59/0.92 | Helped | SPHINX: Structural Prediction using Hypergraph Inference Network (ICML 2025 · 0 citations) |
| Increase transformer blocks from K=2 to K=8 | Synthetic games, NE solving · NE Gap | 0.2239 -> 0.0344; 0.1685 -> 0.0554; 0.1344 -> 0.0484; 0.0796 -> 0.0334 | Helped | NfgTransformer: Equivariant Representation Learning for Normal-form Games (ICLR 2024 · 0 citations) |
| Initialize search with programs sampled from the probabilistic grammar instead of latent samples | Karel and Karel-Hard tasks · episodic return | No clear effect | Reclaiming the Source of Programmatic Policies: Programmatic versus Latent Spaces (ICLR 2024 · 0 citations) | |
| clip the potential difference to [-1, 1] | PinBall · steps to goal | Helped | Goal-Space Planning with Subgoal Models (ICML 2025) | |
| Add the class-dependency escaping strategy, including the conditional mutual information minimization, instead of using only conditional marginal probabilities (DEV1) | 17 MDC benchmark datasets · Hamming Score | Helped | Towards Escaping from Class Dependency Modeling for Multi-Dimensional Classification (ICML 2025) |
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 52 papers; from OpenAlex and the papers' affiliations).
- Microsoft · 2 papers
- Google · 2 papers
- Massachusetts Institute of Technology · 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
- Neuro-symbolic probabilistic layers (120 papers)
- Sparse graphical and tensor models (97 papers)
- Equivariant and disentangled latent representations (90 papers)
- Generative latent models for single-cell data (41 papers)
Elsewhere in the topic
- Generative protein structure design (142 papers)
- Prediction-powered and conformal inference (129 papers)
- Statistical analysis of ML practice (122 papers)
- Structure-based 3D molecule generation (107 papers)
- Flow matching generative modeling (97 papers)
- Machine learning for physical simulation (96 papers)
- Multimodal trajectory prediction (95 papers)
- Counterfactual and statistical explanations (81 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).