ICML 2026 Past Generative models
ICML 2026 Workshop on Structured Probabilistic Inference & Generative Modeling
SPIGM @ ICML
- Submission deadline
- May 9, 2026, 12:00 UTC imported from OpenReview — check the website for extensions
- Submission portal
- OpenReview
- Notes
- Auto-imported from the OpenReview venue record on 2026-06-10 — please verify and enrich (topics are keyword-guessed).
Accepted papers (208)
Fetched from OpenReview (v2) on 2026-06-10.
-
$\psi$DAG: Projected Stochastic Approximation Iteration for Linear DAG Structure Learning
-
A Born Machine Approach to Controllable Text Generation with Language Models
-
A Generative Model for Extremely Sparse Edge-Exchangeable Networks
-
A Mean-Field Framework for Inference-Time Distributional Control of Diffusion Models
-
A Structural View of Query Misspecification in Causal Foundation Models
-
A Tale of Two Temperatures: Simple, Efficient, and Diverse Sampling from Diffusion Language Models
-
A Unified View of Score-Based and Drifting Models
-
ABC: Any-Subset Autoregression via Non-Markovian Diffusion Bridges in Continuous Time and Space
-
Active Flow Expansion for Out-of-Distribution Discovery: from Theory to Molecules
-
Aligning Few-Step Generative Model via Amortizing Sample-Based Variational Inference
-
Alignment-Dependent Inference in Small Language Models via Budgeted Marginalization over Contextual Priors
-
AMIGO: Adapters Meet Information Geometry
-
Amortised Inference through One-Step Implicit Sampling
-
Analytic interdomain memory for efficient online HiPPO-SVGP
-
Anchoring Aleatoric Uncertainty: A Four-Term Decomposition of Predictive Risk at the Bayes-Optimal Predictor
-
Applying Splat Regression Models to Particle Density Control in Radiance Fields
-
Benchmarks as Random Variables—Modeling Overdispersion in LLM Evaluation
-
BIRDGen: Multimodal Conditional Inference of Latent Unbiased Species Distributions
-
Boosting Inference with Guided Reasoning: Stochastic Exploration for Recursive Models
-
Branching Diffusion for Point Processes in Time and Space
-
Breaking the Factorization Barrier in Diffusion Language Models
-
Bridging the Gap Between AI Predictions and Chemical Conventions: Template-Guided Reranking for Accurate Reagent Set Suggestion
-
Calibrating Promptable Concept Segmentation via Paraphrase Consistency
-
Categorical Drifting Models
-
CIRCUS: Circuit Consensus under Uncertainty via Stability Ensembles
-
Compositional Energy-Based Inference-Time Scaling for Multi-Scale Microstructure Generation
-
Compositional Flow Matching with Factored Velocity Fields
-
Conditional Inference Mismatch in Structured Diffusion Language Models
-
Conditional Random Fields for Structured Representation Learning from Pretrained Features
-
Conditional Unbalanced Optimal Transport Maps: An Outlier-Robust Framework for Conditional Generative Modeling
-
Context Over Content: Exposing Evaluation Faking in Automated Judges
-
Context-Aware Neural SDEs for Robust Irregular Time-Series Classification
-
Contour Monte Carlo: Sampling via Energy Level Sets
-
Contrastive Distribution Matching for Amortized Sequential Monte Carlo in Discrete Diffusion
-
Decision-Aware Training for Sample-Based Generative Models
-
Deep Generative Models for Phylogenetic Inference with Complex Evolutionary Processes
-
Deep Heteroskedastic Regression: Post-Hoc Variance Estimation from Latent Representations
-
DELTA-TTS: Adapting Autoregressive Model into a Diffusion Language Model for Text-to-Speech
-
Diagnosing LLM Judge Reliability: Conformal Prediction Sets and Transitivity Violations
-
Diffusion Accelerants: Towards Augmenting Molecular Dynamics with Learned Measure Transport
-
Diffusion Gaussian Processes
-
Direct Flow Neural Processes: Efficient Sampling via Flow Step Amortization
-
Discrete Langevin-Inspired Posterior Sampling
-
DLLM-JEPA: Joint Embedding Predictive Architectures for Masked Diffusion Language Models
-
DODO: Discrete OCR Diffusion Models
-
DualDrift: Combining Forward and Reverse Drifts for One-Step Generative Modeling
-
DUEL: Exact Likelihood for Masked Diffusion via Deterministic Unmasking
-
DVD: Discrete Voxel Diffusion for 3D Generation and Editing
-
Edge-aware FlexAttention Network for Efficient Graph Generation
-
Effective Test-Time Scaling of Discrete Diffusion through Iterative Refinement
-
Efficient One-to-many Domain Translation via Diffusive Entropic Optimal Transport
-
Electrostatic Models for Score Matching
-
End-to-End Context Compression at Scale
-
End-to-End Identifiable and Consistent Recurrent Switching Dynamical Systems
-
Enhanced Diffusion Sampling: Efficient Rare Event Sampling and Free Energy Calculation with Diffusion Models
-
Evolutionary Curriculum Learning for Biological Sequence Modeling
-
Exact Posterior Score Estimation for Solving Linear Inverse Problems
-
Expanding Flow Maps
-
Extracting Local Manifold Geometry from Pretrained Diffusion Models in One Inverse Step
-
Factored Score Matching on Graphical Models: Exact Computation on Trees and Convergent Approximation on Loopy Graphs
-
FairOpt-PFN: Amortized Counterfactual Fairness with Optimal Fair Targets
-
Fast-dLLM++: Fr\'{e}chet Profile Decoding for Faster Diffusion LLM Inference
-
Faster Inference for Conditional Masked Diffusion Language Models by Knowledge Distillation of Guidance and Trajectory
-
Federated Learning with Energy-Based Structured Probabilistic Inference
-
Federated Sampling of Molecular Conformers via Compositional Flows
-
Few-Step Boltzmann Generators via Scalable Likelihood Flow Maps
-
Finetuning Generative Models to Match Feature Distributions
-
Fisher-constrained flow matching for transferable free energy estimation
-
Fixed-Point Distillation of Flow Matching Models
-
Fixed-Point Masked Generative Modeling
-
Flash-SD-KDE: Accelerating SD-KDE with Tensor Cores
-
Flow Map Denoisers: Traversing the Distortion-Perception Plane for Inverse Problems
-
Flow Matching for Reaction Pathway Generation
-
Flow Matching on General Manifolds via Pulling Back Geodesic Convex Latent Manifolds
-
FM-DeepRV: Deep Learning for Bayesian Inference with Flow Matching
-
Forecasting Motion in the Wild
-
Frequency-Forcing: From Scaling-as-Time to Soft Frequency Guidance
-
From Fisher--Rao Simplex Flows to Canonical Jump Generators: A $\Gamma$-Convergence Theory of Discrete Flow Matching
-
Frontier Language Models Struggle to Copy: Text Can Be Better Viewed in 2D
-
GAP3D: Generative Alignment of VLM Latents to Patch-Level Embeddings for 3D Generation
-
Gaussian Particle Flows for Unsupervised Topology Optimization
-
Gene-Embedding Perturbation Operators for Zero-Shot and Transferable Prediction of Transcriptional Responses
-
Generalised Latent Slice Sampling
-
Generative Modeling via Kernelized Stochastic Interpolants
-
GRIFDIR: Graph Resolution-Invariant FEM Diffusion Models in Function Spaces over Irregular Domains
-
Hacking Generative Perplexity: Why Unconditional Text Evaluation Needs Distributional Metrics
-
Holistic Latent Diffusion Acceleration: Unifying Spatial, Temporal, and Architectural Efficiency
-
How Deep Are Deep GPs, Really? A Sharp Threshold and a Non-Gaussian Limit for Compositional GPs
-
How to Spend Your Oracle Budget: Practical Guidance for Protein Structure Foundation Models
-
How to Train Your Latent Diffusion Language Model Jointly With the Latent Space
-
Hyperbolic Latent Geometry for Tree-Structured Prototype Networks: A Local-vs-Global Trade-off
-
Implicit Neural Representations of Individual Behavior
-
Improving Conformal Prediction Sets Through Semantic Neighborhood Diffusion
-
Integrating Causal DAGs in Deep RL: Activating Minimal Markovian States with Multi-Order Exposure
-
Inter-Trajectory Importance Sampling Improves Diffusion Samplers
-
Internal Data Repetition Destroys Language Models
-
Inverse problems with diffusion models: MAP estimation via mode-seeking loss
-
Inverse-Confidence Sampling for Continuous Diffusion Language Models
-
Inverting Foundation Models of Brain Function with Simulation-Based Inference
-
Irregularities of Latent Space Geometry in Diffusion Models
-
Isokinetic Flow Matching for Pathwise Straightening
-
Just Add More Capacitors: Eliminating Flux Leakage in Electrostatic Field Matching
-
Kernel-Gradient Drifting Models
-
LangFlow: Continuous Diffusion Rivals Discrete in Language Modeling
-
Language Models Need Sleep
-
Latent-Augmented Discrete Diffusion Models
-
Learn from Your Mistakes: Self-Correcting Masked Diffusion Models
-
Learned Relay Representations for Forward-Thinking Discrete Diffusion Models
-
Learning Adapter Rank via Symmetry Breaking
-
Learning Manifold Data with Flow Matching
-
Learning path splines via Acceleration Matching
-
Learning Shortest Paths with Generative Flow Networks
-
Learning to Shift Numeric Predictive Densities for Uncertainty-Aware LLM Agents
-
Leveraging Generative Mode-Seeking for Precision Matrix Estimation
-
Limit Order Book Forecasting with Conditional Diffusion Models
-
MCD-RRG: Time-Varying Multimodal Fusion and Residual Retrieval Guidance for Conditional Diffusion
-
Measuring and Reducing Train--Inference Mismatch in Discrete Diffusion Language Models
-
Midpoint Generative Models
-
MIRROR: Multisensory Implicit Rejection-sampled RObotic policy
-
MIST: Mutual Information Estimation via Supervised Training
-
Model-Free Assessment of Simulator Fidelity via Quantile Curves
-
Multilingual Synthetic Scanpaths: Cross-Language Generalization for Gaze Generation
-
Neuro-Symbolic ODE Discovery with Latent Grammar Flow
-
Noise Scheduling as Information-Guided Allocation in Diffusion Training
-
Nonparametric Distribution Matching for Self-Supervised Whole-Slide Image Condensation
-
Normalizing Trajectory Models
-
On Calibration of Modern Language Models
-
On the Difficulty of Feature Unlearning in Tabular Diffusion Models
-
ORBIT: Counterfactual Proposal Inference for Prompt-Free 3D Brain Tumor Segmentation
-
Order-Agnostic Decoding for Sample-Efficient RNA Inverse Folding
-
OrthoBO: Orthogonal Bayesian Hyperparameter Optimization
-
PairIT: Autoregressive Transformers for Low-Data Molecule Optimization
-
Parallel Tempering Initial Sampling in Inference-Time Reward Alignment
-
Path-independent Flow Matching for Multi-parameter Generative Dynamics
-
Perfect Recall, Parallel Efficiency: Interleaved DeepSeek Sparse Attention for Million-Token-Context Decoding
-
Pi-E-Flow: Uncertainty-Guided Flow Distillation for Autoregressive Video Generation
-
Plan, Don’t Pose: Long Composite Motion Generation with Text-Aligned BFM
-
Position: Benchmark Method-Comparisons Are Posterior Identifiability Problems
-
Position: Multi-Agent LLM Simulation as Approximate Posterior Inference Demands a Probabilistic Calibration Standard
-
Prior-Informed Flow Matching for Graph Reconstruction
-
PRISM-SLAM: Probabilistic Ray-Grounded Inference for Scale-aware Metric SLAM
-
Probabilistic Chain-of-Thought: Sequential Bayesian Inference over Latent Reasoning Correctness
-
Probabilistic Sequence Generation Guided by Intensity-Duration Extreme Profiles
-
Provably Stable Neural Dynamics via Koopman Operator Certificates
-
Proximal Policy Optimization for Amortized Discrete Sampling
-
Random-Projection Tree Stein Variational Gradient Descent
-
Rao-Blackwellized Score Matching on Manifolds
-
RDDMPI: Residual Denoising Diffusion Model for Probabilistic Multivariate Time Series Imputation
-
Re-evaluating Confidence Remasking in Masked Diffusion Language Models
-
Readout Times Are Not Solver Nodes: A Two-Mesh API for Generative ODE Surrogates
-
ReCache: Learning Budget-Aware Caching Schedules for Diffusion Models via REINFORCE
-
Reconsidering Positional Supervision in Masked Diffusion Language Model Training
-
Registers Matter for Pixel-space Diffusion Transformers
-
Residual-Space Evolutionary Optimization via Flow-based Generative Models
-
Reward Score Matching: Unifying Reward-based Fine-tuning for Flow and Diffusion Models
-
Reward-Aligning Few-Step Flow Models with Integrated Regularizers
-
Reweighted ALPS: Non-Asymptotic Guarantees for Multimodal Sampling with Warm Starts
-
SALSA: State Augmentation via Learned Selective Attention
-
SASC: Soft-Averaged Self-Consistency to Improve Chain-of-Thought Reasoning in Instruct-LLMs
-
Savitar: Curve-Aware Interaction-Structured Kernels for Low-Budget Bayesian Optimization in Rare-Winner Combinatorial Spaces
-
Scalable Bayesian Monte Carlo: fast uncertainty estimation beyond deep ensembles
-
Scalable Deep Basis Kernel Gaussian Processes
-
Scalable Differentially Private Data Compression via Diffusion and Stochastic Codes
-
Scalable Inference-Time Steering in Molecular Design with Multimodal Meta Flow Maps
-
Scaling with Recursion in Masked Discrete Diffusion Models
-
Self-Supervised Variational Priors for Robust Bayesian Inference
-
Signal from Structure: Exploiting Submodular Upper Bounds in Generative Flow Networks
-
Single-Step Initialization for Exploratory Parallel Rollouts in Diffusion LLMs
-
Size- and Dispersion-Corrected Two-Level Softmax Sampling
-
Solving Integer Linear Programming with Parallel Tempering
-
SpatialNP: Gridded Transformer Neural Processes for Probabilistic Spatial Proteomics in Multiplexed Tissue Imaging
-
Stable and Near-Reversible Diffusion ODE Solvers for Image Editing
-
STARFlow2: Bridging Language Models and Normalizing Flows for Unified Multimodal Generation
-
STARS: Synchronous Token Alignment for Robust Supervision in Large Language Models
-
Stop the Sampler! Classifier-Based Adaptive Stopping for Sampling Kernels
-
Strong Stochastic Flow Maps
-
Structural Support Certificates for Graph-Query Inference in Mechanism Posteriors
-
Structured Coupling for Flow Matching
-
Structured Inference with Large Language Gibbs
-
Structuring The Future: Diffusion LLM Speculative Decoding via Calibrated Draft Graphs
-
Synthesizability-Aware Materials Generation with Target Properties via Reinforcement Learning
-
Systematic Study of Grid Adaptation Strategies in Kolmogorov–Arnold Networks
-
TAPS: Task Aware Proposal Distributions for Speculative Sampling
-
Tensor-Train Joint Modeling for Few-Step Discrete Diffusion
-
The Confidence Shortcut: A Reasoning Failure Mode of Masked Diffusion Models
-
The Model Knows, the Decoder Finds: Future Value Guided Particle Power Sampling
-
TILT: Test-Time Reward Alignment via Distribution Tilting for Compositional Generation
-
Time-Annealed Perturbation Sampling: Diverse Generation for Diffusion Language Models
-
Time-Correlated Video Bridge Matching
-
Topological Control of Optimization Dynamics on Evolving Manifolds
-
Towards Closing the Autoregressive Gap in Language Modeling via Entropy-Gated Continuous Bitstream Diffusion
-
TUBE: Tangent Upper Bound on Evidence for Discrete Diffusion Language Models
-
U-Former ODE: Fast Probabilistic Forecasting of Irregular Time Series
-
Uncertainty Estimation for Molecular Diffusion Models
-
Uncertainty Quantification for LLM Agents via Semantic Abstraction Trajectories
-
Understanding and Accelerating the Training of Masked Diffusion Language Models
-
Uniform Diffusion Models Revisited: Leave-One-Out Denoiser and Absorbing State Reformulation
-
Unlocking the Duality between Flow and Field Matching
-
Variance Reduction for Expectations with Diffusion Teachers
-
Variance-Tilted Diffusion Models for Diverse Sampling
-
WarmPrior: Straightening Flow-Matching Policies with Temporal Priors
-
Wasserstein Gradient Flows and Forward-Only Diffusion Are Not Enough for Multimodal Sampling
-
Wasserstein Residuals: Learning Gradient Flows from Population Dynamics
-
When are likely answers right? On Sequence Probability and Correctness in LLMs
-
When Does a Low-Rank Bayesian Neural Network Certify Its Deterministic Center?
-
When Inference-Time Reward Steering Hacks the Reward
-
Your Autoregressive Model Already Reveals the Causal Graph
-
Your GFlowNet Secretly Learns an Optimal Transport Plan