ICLR 2026 accepted papers on Probabilistic and scientific modeling
315 of the 5,358 papers accepted at ICLR 2026 (5.9%) are on Probabilistic and scientific modeling, compared with 7.8% at ICLR 2025. They fall into 8 topics.
See ICLR 2026 in the interactive map
Working on something in this area? Describe your idea in scime atlas to see the closest accepted papers by meaning and how crowded the spot has become.
Topics and papers
Molecules and proteins · 83 papers
- SAIR: Enabling Deep Learning for Protein-Ligand Interactions with a Synthetic Structural Dataset (11 citations)
- SimpleFold: Folding Proteins is Simpler than You Think (3 citations)
- A New Paradigm for Genome-wide DNA Methylation Prediction Without Methylation Input (2 citations)
- Learning from the Electronic Structure of Molecules across the Periodic Table (2 citations)
- PepBenchmark: A Standardized Benchmark for Peptide Machine Learning (2 citations)
- Protein Structure Tokenization via Geometric Byte Pair Encoding (2 citations)
- Exploring Synthesizable Chemical Space with Iterative Pathway Refinements (1 citation)
- Graph Diffusion Transformers are In-Context Molecular Designers (1 citation)
- I Predict Therefore I Am: Is Next Token Prediction Enough to Learn Human-Interpretable Concepts from Data? (1 citation)
- Interpolation-Based Conditioning of Flow Matching Models for Bioisosteric Ligand Design (1 citation)
- Out of the Shadows: Exploring a Latent Space for Neural Network Verification (1 citation)
- Property-Driven Protein Inverse Folding with Multi-Objective Preference Alignment (1 citation)
and 71 more in the interactive map.
Variational and generative sampling · 65 papers
- AlphaFlow: Understanding and Improving MeanFlow Models (1 citation)
- A Statistical Benchmark for Diffusion Posterior Sampling Algorithms
- Accelerated Parallel Tempering via Neural Transports
- AlignFlow: Improving Flow-based Generative Models with Semi-Discrete Optimal Transport
- Bi-Lipschitz Autoencoder With Injectivity Guarantee
- Branched Schrödinger Bridge Matching
- Carré du champ flow matching: better quality-generalisation tradeoff in generative models
- Contact Wasserstein Geodesics for Non-Conservative Schrödinger Bridges
- Count Bridges enable Modeling and Deconvolving Transcriptomic Data
- Curse of Slicing: Why Sliced Mutual Information is a Deceptive Measure of Statistical Dependence
- Discrete Adjoint Matching
- Discrete Variational Autoencoding via Policy Search
and 53 more in the interactive map.
Causal discovery · 38 papers
- Addressing divergent representations from causal interventions on neural networks
- Beyond DAGs: A Latent Partial Causal Model for Multimodal Learning
- CaTs and DAGs: Integrating Directed Acyclic Graphs with Transformers for Causally Constrained Predictions
- Causal Discovery via Quantile Partial Effect
- Causal Structure Learning in Hawkes Processes with Complex Latent Confounder Networks
- Characterization and Learning of Causal Graphs with Latent Confounders and Post-treatment Selection from Interventional Data
- Coarse-to-Fine Learning of Dynamic Causal Structures
- Distributional Equivalence in Linear Non-Gaussian Latent-Variable Cyclic Causal Models: Characterization and Learning
- Diverse Dictionary Learning
- Efficient Ensemble Conditional Independence Test Framework for Causal Discovery
- Embracing Discrete Search: A Reasonable Approach to Causal Structure Learning
- Foundation Models for Causal Inference via Prior-Data Fitted Networks
and 26 more in the interactive map.
Differential equations and PDEs · 36 papers
- Fast training of accurate physics-informed neural networks without gradient descent (2 citations)
- A Two-Phase Deep Learning Framework for Adaptive Time-Stepping in High-Speed Flow Modeling
- Adaptive Mamba Neural Operators
- CFO: Learning Continuous-Time PDE Dynamics via Flow-Matched Neural Operators
- ComPhy: Composing Physical Models with end-to-end Alignment
- Deep Learning for Subspace Regression
- Enhancing Stability of Physics-Informed Neural Network Training Through Saddle-Point Reformulation
- Extending Fourier Neural Operators for Modeling Parameterized and Coupled PDEs
- Generalized Spherical Neural Operators: Green’s Function Formulation
- Iterative Training of Physics-Informed Neural Networks with Fourier-enhanced Features
- LD-EnSF: Synergizing Latent Dynamics with Ensemble Score Filters for Fast Data Assimilation with Sparse Observations
- Learning Data-Efficient and Generalizable Neural Operators via Fundamental Physics Knowledge
and 24 more in the interactive map.
Dynamical systems · 33 papers
- Coupled Transformer Autoencoder for Disentangling Multi-Region Neural Latent Dynamics (1 citation)
- From movement to cognitive maps: recurrent neural networks reveal how locomotor development shapes hippocampal spatial coding (1 citation)
- Self-Supervised Evolution Operator Learning for High-Dimensional Dynamical Systems (1 citation)
- A Spectral-Grassmann Wasserstein metric for operator representations of dynamical systems
- Discovering heterogeneous synaptic plasticity rules via large-scale neural evolution
- Emergence of Spatial Representation in an Actor-Critic Agent with Hippocampus-Inspired Sequence Generator
- Feedback-driven recurrent quantum neural network universality
- GenCP: Towards Generative Modeling Paradigm of Coupled physics
- Homeostatic Adaptation of Optimal Population Codes under Metabolic Stress
- Identifiability Challenges in Sparse Linear Ordinary Differential Equations
- InputDSA: Demixing, then comparing recurrent and externally driven dynamics
- Learning Collective Variables from BioEmu with Time-Lagged Generation
and 21 more in the interactive map.
Uncertainty and Gaussian processes · 27 papers
- Compositional amortized inference for large-scale hierarchical Bayesian models (1 citation)
- Multifidelity Simulation-based Inference for Computationally Expensive Simulators (1 citation)
- Amortising Inference and Meta-Learning Priors in Neural Networks
- Bayesian Neural Networks for Functional ANOVA Model
- Contextual Similarity Distillation: Ensemble Uncertainties with a Single Model
- CUPID: A Plug-in Framework for Joint Aleatoric and Epistemic Uncertainty Estimation with a Single Model
- Diffusion Bridge Variational Inference for Deep Gaussian Processes
- Efficient Credal Prediction through Decalibration
- Epistemic Uncertainty Quantification To Improve Decisions From Black-Box Models
- Geometric Autoencoder Priors for Bayesian Inversion: Learn First Observe Later
- Optimizing Data Augmentation through Bayesian Model Selection
- Robust Amortized Bayesian Inference with Self-Consistency Losses on Unlabeled Data
and 15 more in the interactive map.
Bayesian optimization · 17 papers
- A Genetic Algorithm for Navigating Synthesizable Molecular Spaces (1 citation)
- Adaptive Acquisition Selection for Bayesian Optimization with Large Language Models
- BioBO: Biology-informed Bayesian Optimization for Perturbation Design
- Combinatorial Bandit Bayesian Optimization for Tensor Outputs
- Discount Model Search for Quality Diversity Optimization in High-Dimensional Measure Spaces
- From Sorting Algorithms to Scalable Kernels: Bayesian Optimization in High-Dimensional Permutation Spaces
- GIT-BO: High-Dimensional Bayesian Optimization with Tabular Foundation Models
- Local Entropy Search over Descent Sequences for Bayesian Optimization
- Symmetry-Aware Bayesian Optimization via Max Kernels
- Task-free Adaptive Meta Black-box Optimization
- Thompson Sampling via Fine-Tuning of LLMs
- Generative Bayesian Optimization: Generative Models as Acquisition Functions
and 5 more in the interactive map.
Treatment effects · 16 papers
- A Relative Error-Based Evaluation Framework of Heterogeneous Treatment Effect Estimators
- Direct Doubly Robust Estimation of Conditional Quantile Contrasts
- Good allocations from bad estimates
- IGC-Net for conditional average potential outcome estimation over time
- Journey to the Centre of Cluster: Harnessing Interior Nodes for A/B Testing under Network Interference
- Learning Exposure Mapping Functions for Inferring Heterogeneous Peer Effects
- Overlap-Adaptive Regularization for Conditional Average Treatment Effect Estimation
- Overlap-weighted orthogonal meta-learner for treatment effect estimation over time
- Stochastic Neural Networks for Causal Inference with Missing Confounders
- SurvHTE-Bench: A Benchmark for Heterogeneous Treatment Effect Estimation in Survival Analysis
- ActiveCQ: Active Estimation of Causal Quantities
- Adjusting Prediction Model Through Wasserstein Geodesic for Causal Inference
and 4 more in the interactive map.