ICML 2025 accepted papers on Probabilistic and scientific modeling
341 of the 3,342 papers accepted at ICML 2025 (10.2%) are on Probabilistic and scientific modeling, compared with 12.3% at ICML 2024. They fall into 8 topics.
See ICML 2025 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 · 104 papers
- Learning Smooth and Expressive Interatomic Potentials for Physical Property Prediction (33 citations)
- P(all-atom) Is Unlocking New Path For Protein Design (11 citations)
- DiffMS: Diffusion Generation of Molecules Conditioned on Mass Spectra (10 citations)
- Sidechain conditioning and modeling for full-atom protein sequence design with FAMPNN (6 citations)
- Fixing the Double Penalty in Data-Driven Weather Forecasting Through a Modified Spherical Harmonic Loss Function (5 citations)
- Inverse problems with experiment-guided AlphaFold (5 citations)
- GenMol: A Drug Discovery Generalist with Discrete Diffusion (4 citations)
- The dark side of the forces: assessing non-conservative force models for atomistic machine learning (4 citations)
- Boosting Protein Graph Representations through Static-Dynamic Fusion (3 citations)
- Empower Structure-Based Molecule Optimization with Gradient Guided Bayesian Flow Networks (3 citations)
- LDMol: A Text-to-Molecule Diffusion Model with Structurally Informative Latent Space Surpasses AR Models (3 citations)
- CellFlux: Simulating Cellular Morphology Changes via Flow Matching (2 citations)
and 92 more in the interactive map.
Variational and generative sampling · 70 papers
- Sampling Binary Data by Denoising through Score Functions (2 citations)
- A Theoretical Framework For Overfitting In Energy-based Modeling (1 citation)
- Latent Thought Models with Variational Bayes Inference-Time Computation (1 citation)
- MITIGATING OVER-EXPLORATION IN LATENT SPACE OPTIMIZATION USING LES (1 citation)
- Monte-Carlo Tree Search with Uncertainty Propagation via Optimal Transport (1 citation)
- Scalable Equilibrium Sampling with Sequential Boltzmann Generators (1 citation)
- Synonymous Variational Inference for Perceptual Image Compression (1 citation)
- A Likelihood Based Approach to Distribution Regression Using Conditional Deep Generative Models
- A Sample Efficient Conditional Independence Test in the Presence of Discretization
- A Theory for Conditional Generative Modeling on Multiple Data Sources
- A Unified View on Learning Unnormalized Distributions via Noise-Contrastive Estimation
- Aggregation of Dependent Expert Distributions in Multimodal Variational Autoencoders
and 58 more in the interactive map.
Causal discovery · 41 papers
- Teaching Transformers Causal Reasoning through Axiomatic Training (2 citations)
- Discovering Latent Causal Graphs from Spatiotemporal Data (1 citation)
- Identifying biological perturbation targets through causal differential networks (1 citation)
- MIB: A Mechanistic Interpretability Benchmark (1 citation)
- A Causal World Model Underlying Next Token Prediction: Exploring GPT in a Controlled Environment
- A New Approach to Backtracking Counterfactual Explanations: A Unified Causal Framework for Efficient Model Interpretability
- An Efficient Search-and-Score Algorithm for Ancestral Graphs using Multivariate Information Scores for Complex Non-linear and Categorical Data
- Arrow: Accelerator for Time Series Causal Discovery with Time Weaving
- Causal Abstraction Learning based on the Semantic Embedding Principle
- Counterfactual Voting Adjustment for Quality Assessment and Fairer Voting in Online Platforms with Helpfulness Evaluation
- Distinguishing Cause from Effect with Causal Velocity Models
- E-LDA: Toward Interpretable LDA Topic Models with Strong Guarantees in Logarithmic Parallel Time
and 29 more in the interactive map.
Differential equations and PDEs · 31 papers
- Linearization Turns Neural Operators into Function-Valued Gaussian Processes (2 citations)
- Toward Efficient Kernel-Based Solvers for Nonlinear PDEs (2 citations)
- Unisolver: PDE-Conditional Transformers Towards Universal Neural PDE Solvers (2 citations)
- Interpolating Neural Network-Tensor Decomposition (INN-TD): a scalable and interpretable approach for large-scale physics-based problems (1 citation)
- Learn Singularly Perturbed Solutions via Homotopy Dynamics (1 citation)
- OmniArch: Building Foundation Model for Scientific Computing (1 citation)
- Sub-Sequential Physics-Informed Learning with State Space Model (1 citation)
- Transolver++: An Accurate Neural Solver for PDEs on Million-Scale Geometries (1 citation)
- Active Learning with Selective Time-Step Acquisition for PDEs
- Calibrated Physics-Informed Uncertainty Quantification
- Closed-form Solutions: A New Perspective on Solving Differential Equations
- DISCO: learning to DISCover an evolution Operator for multi-physics-agnostic prediction
and 19 more in the interactive map.
Uncertainty and Gaussian processes · 30 papers
- Addressing Misspecification in Simulation-based Inference through Data-driven Calibration (3 citations)
- BARNN: A Bayesian Autoregressive and Recurrent Neural Network (1 citation)
- Bayesian Inference for Correlated Human Experts and Classifiers
- Bayesian Neural Scaling Law Extrapolation with Prior-Data Fitted Networks
- Can Transformers Learn Full Bayesian Inference in Context?
- Can We Predict Performance of Large Models across Vision-Language Tasks?
- Confidential Guardian: Cryptographically Prohibiting the Abuse of Model Abstention
- Conformal Prediction as Bayesian Quadrature
- Decision Making under the Exponential Family: Distributionally Robust Optimisation with Bayesian Ambiguity Sets
- Deep Bayesian Filter for Bayes-Faithful Data Assimilation
- Fully Heteroscedastic Count Regression with Deep Double Poisson Networks
- Improving Generalization with Flat Hilbert Bayesian Inference
and 18 more in the interactive map.
Dynamical systems · 28 papers
- Flow-field inference from neural data using deep recurrent networks (9 citations)
- Deep Neural Cellular Potts Models (1 citation)
- Universal Approximation of Mean-Field Models via Transformers (1 citation)
- A Generalizable Physics-Enhanced State Space Model for Long-Term Dynamics Forecasting in Complex Environments
- Accurate Identification of Communication Between Multiple Interacting Neural Populations
- CogReact: A Reinforced Framework to Model Human Cognitive Reaction Modulated by Dynamic Intervention
- Discovering Physics Laws of Dynamical Systems via Invariant Function Learning
- Dynamical Modeling of Behaviorally Relevant Spatiotemporal Patterns in Neural Imaging Data
- Exploring Invariance in Images through One-way Wave Equations
- Guided Structural Inference: Leveraging Priors with Soft Gating Mechanisms
- Learning Dynamics under Environmental Constraints via Measurement-Induced Bundle Structures
- Learning Time-Varying Multi-Region Brain Communications via Scalable Markovian Gaussian Processes
and 16 more in the interactive map.
Treatment effects · 25 papers
- On Efficient Estimation of Distributional Treatment Effects under Covariate-Adaptive Randomization (2 citations)
- Integer Programming for Generalized Causal Bootstrap Designs (1 citation)
- A Meta-learner for Heterogeneous Effects in Difference-in-Differences
- Adjustment for Confounding using Pre-Trained Representations
- AutoCATE: End-to-End, Automated Treatment Effect Estimation
- Causal Effect Identification in lvLiNGAM from Higher-Order Cumulants
- Double Machine Learning for Causal Inference under Shared-State Interference
- Doubly Protected Estimation for Survival Outcomes Utilizing External Controls for Randomized Clinical Trials
- Enhancing Treatment Effect Estimation via Active Learning: A Counterfactual Covering Perspective
- Falsification of Unconfoundedness by Testing Independence of Causal Mechanisms
- Generalized Random Forests Using Fixed-Point Trees
- Heterogeneous Treatment Effect in Time-to-Event Outcomes: Harnessing Censored Data with Recursively Imputed Trees
and 13 more in the interactive map.
Bayesian optimization · 12 papers
- FunBO: Discovering Acquisition Functions for Bayesian Optimization with FunSearch (1 citation)
- Multi-Objective Causal Bayesian Optimization (1 citation)
- A Unified Framework for Entropy Search and Expected Improvement in Bayesian Optimization
- Bayesian Optimization from Human Feedback: Near-Optimal Regret Bounds
- Density Ratio Estimation-based Bayesian Optimization with Semi-Supervised Learning
- Multi-Timescale Dynamics Model Bayesian Optimization for Plasma Stabilization in Tokamaks
- Optimization Proxies using Limited Labeled Data and Training Time -- A Semi-Supervised Bayesian Neural Network Approach
- Towards Universal Offline Black-Box Optimization via Learning Language Model Embeddings
- Understanding High-Dimensional Bayesian Optimization
- BARK: A Fully Bayesian Tree Kernel for Black-box Optimization
- Natural Perturbations for Black-box Training of Neural Networks by Zeroth-Order Optimization
- Posterior Inference with Diffusion Models for High-dimensional Black-box Optimization