ICLR 2025 Past Generative modelsTheory
ICLR 2025 Workshop on Deep Generative Model in Machine Learning: Theory, Principle and Efficacy
ICLR 2025 DeLTa Workshop
- Submission deadline
- Feb 12, 2025, 00:59 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 (125)
Fetched from OpenReview (v2) on 2026-06-10.
-
A Reversible Solver for Diffusion SDEs
-
A Simple Model of Inference Scaling Laws
-
A Theory for Conditional Generative Modeling on Multiple Data Sources
-
A Unified Diffusion Bridge Framework via Stochastic Optimal Control
-
ADAPTIVE HETEROGENEOUS GRAPH REPRESENTATION LEARNING USING KNN-AUGMENTED GRAPH MAMBA NETWORKS (KA-GMN)
-
An Improved Sample Complexity for Rank-1 Matrix Sensing
-
AtropDiff: Data-Scarce Atropisomer Generation via Multi-Task Pretrained Classifier-Guided Diffusion
-
Attention Scheme Inspired Softmax Regression
-
Balanced Latent Space of Diffusion Models for Counterfactual Generation
-
Breaking the Likelihood--Quality Trade-off in Diffusion Models by Merging Pretrained Experts
-
BridgeVoC: Insights into Using Schrödinger Bridge for Neural Vocoders
-
Building A Unified AI-centric Language System: analysis, framework and future work
-
Can Diffusion Models Learn Hidden Inter-Feature Rules Behind Images?
-
Causal Representation Learning and Inference via Mixture-Based Priors
-
Cellular-Guided Graph Generative Model
-
Chimera: State Space Models Beyond Sequences
-
CoDe: Blockwise Control for Denoising Diffusion Models
-
Computational Limits of Low-Rank Adaptation (LoRA) Fine-Tuning for Transformer Models
-
DDPM Score Matching Is Asymptotically Efficient
-
DEEP CLUSTERING USING ADVERSARIAL NET BASED CLUSTERING LOSS
-
Demystifying Long Chain-of-Thought Reasoning in LLMs
-
Demystifying the Token Dynamics of Deep Selective State Space Models
-
Design Editing for Offline Model-based Optimization
-
Designing Parameter and Compute Efficient Diffusion Transformers using Distillation
-
Diffusion Models Do Not Implicitly Learn Conditional Independence
-
DIFFUSION MODELS LEARN LOW-DIMENSIONAL DISTRIBUTIONS VIA SUBSPACE CLUSTERING
-
Diffusion-Based Planning for Autonomous Driving with Flexible Guidance
-
DIME: Deterministic Information Maximizing Autoencoder
-
Distance-Based Tree-Sliced Wasserstein Distance
-
DOSE3 : Diffusion-based Out-of-distribution detection on SE(3) trajectories
-
Edge-preserving noise for diffusion models
-
EDM2+: Exploring Efficient Diffusion Model Architectures for Visual Generation
-
Efficient Consistency Model Training for Policy Distillation in Reinforcement Learning
-
Efficient Distributed Optimization under Heavy-Tailed Noise
-
Efficient Knowledge Distillation via Curriculum Extraction
-
Efficient Molecular Conformer Generation with SO(3) Averaged Flow-Matching and Reflow
-
Efficient Multi-View Driving Scenes Generation Based on Video Diffusion Transformer
-
Entropic Time Schedulers for Generative Diffusion Models
-
Flow Along the K-Amplitude for Generative Modeling
-
Flow Matching Neural Processes
-
Flows don't cross in high dimension
-
Fourier Head: Helping Large Language Models Learn Complex Probability Distributions
-
Frame Generation in Hilbert Space: Generative Interpolation of Measurement Data for Quantum Parameter Adaptation
-
FullDiffusion: Diffusion Models Without Time Truncation
-
Fundamental Limits of Prompt Tuning Transformers: Universality, Capacity and Efficiency
-
Gauge Flow Matching for Efficient Constrained Generative Modeling over General Convex Set
-
Graph Discrete Diffusion: a Spectral Study
-
GRAPH GENERATIVE PRE-TRAINED TRANSFORMER
-
Graph transformers express monadic second-order logic
-
Guiding Time-Varying Generative Models with Natural Gradients on Exponential Family Manifold
-
Gumbel-Softmax Score and Flow Matching for Discrete Biological Sequence Generation
-
Hidden in the Noise: Two-Stage Robust Watermarking for Images
-
Hiding and Recovering Knowledge in Text-to-Image Diffusion Models via Learnable Prompts
-
High-Order Matching for One-Step Shortcut Diffusion Models
-
How Compositional Generalization and Creativity Improve as Diffusion Models are Trained
-
How Well Does Your Tabular Generator Learn the Structure of Tabular Data?
-
Identifiable Object Representations under Spatial Ambiguities
-
Identifying metric structures of deep latent variable models
-
Image Interpolation with Score-based Riemannian Metrics of Diffusion Models
-
Image-Alchemy : Advancing Subject Fidelity in Personalized Text-to-Image Generation
-
Implicit Bayesian Inference is An Insufficient Explanation of Language Model Behaviour in Compositional Tasks
-
Improved Techniques for Training Smaller and Faster Stable Diffusion
-
Improving Single Noise Level Denoising Samplers with Restricted Gaussian Oracles
-
Improving Vector-Quantized Image Modeling with Latent Consistency-Matching Diffusion
-
INFO-SEDD: Continuous Time Markov Chains as Scalable Information Metrics Estimators
-
Interleaved Gibbs Diffusion for Constrained Generation
-
LaM-SLidE: Latent Space Modeling of Spatial Dynamical Systems via Linked Entities
-
LapLoss: Laplacian Pyramid-based Multiscale Loss for Image Translation
-
Large Language Diffusion Models
-
Latent Diffusion U-Net Representations Contain Positional Embeddings and Anomalies
-
LEARNING STRAIGHT FLOWS BY LEARNING CURVED INTERPOLANTS
-
Leveraging shared feature representation in cross-domain alignment of decision thresholds for electronic health records data.
-
Mamba State-Space Models Are Lyapunov-Stable Learners
-
Masked Generative Nested Transformers with Decode Time Scaling
-
MCM: Multi-layer Concept Map for Efficient Concept Learning from Masked Images
-
Measuring Semantic Information Production in Generative Diffusion Models
-
Mixture-of-Mamba: Enhancing Multi-Modal State-Space Models with Modality-Aware Sparsity
-
Mixture-of-Transformers: A Sparse and Scalable Architecture for Multi-Modal Foundation Models
-
Multi-view Geometry-Aware Diffusion Transformer for Indoor Novel View Synthesis
-
Neural Genetic Search in Discrete Spaces
-
Nonparametric Distributional Black-box Optimization via Diffusion Process
-
On Distilling Generator Matching Models
-
On the Cone Effect in the Learning Dynamics
-
On the Power of Context Enhanced Learning in LLMs
-
On the Query Complexity of Verifier-Assisted Language Generation
-
Optimizing GPT for Video Understanding: Zero-Shot Performance and Prompt Engineering
-
Outsourced diffusion sampling: Efficient posterior inference in latent spaces of generative models
-
PAC Privacy Preserving Diffusion Models
-
Path Planning for Masked Diffusion Models with Applications to Biological Sequence Generation
-
Phase-aware Training Schedule Simplifies Learning in Flow-Based Generative Models
-
PHYSICS-INFORMED GENERATIVE APPROACHES FOR WIRELESS CHANNEL MODELING
-
Probability-Flow ODE in Infinite-Dimensional Function Spaces
-
Provable Maximum Entropy Manifold Exploration via Diffusion Models
-
Remasking Discrete Diffusion Models with Inference-Time Scaling
-
Revisiting Noise Schedule Design for Diffusion Training
-
Reward-Guided Diffusion Model for Data-Driven Black-Box Design Optimization
-
RFMI: Estimating Mutual Information on Rectified Flow for Text-to-Image Alignment
-
Score as Action: Fine-Tuning Diffusion Generative Models by Continuous-time Reinforcement Learning
-
SDE Matching: Scalable and Simulation-Free Training of Latent Stochastic Differential Equations
-
Shaping Inductive Bias in Diffusion Models through Frequency-Based Noise Control
-
Solving Bayesian inverse problems with diffusion priors and off-policy RL
-
SpecSTG: A Fast Spectral Diffusion Framework for Probabilistic Spatio-Temporal Traffic Forecasting
-
Spherical Tree-Sliced Wasserstein Distance
-
Stable Consistency Tuning: Understanding and Improving Consistency Models
-
Statistical Foundations of Conditional Diffusion Transformers
-
StochSync: Stochastic Diffusion Synchronization for Image Generation in Arbitrary Spaces
-
Symmetry Is All You Need: Image Generation Using Pre-trained Deep Diffusion Probabilistic Models
-
Symmetry-Preserving Diffusion Models via Target Symmetrization
-
TASKD-LLM: Task-Aware Selective Knowledge Distillation for LLMs
-
The Diffusion Duality
-
The Space Between: On Folding, Symmetries and Sampling
-
Towards Black-Box Membership Inference Attack for Diffusion Models
-
Towards Training One-Step Diffusion Models Without Distillation
-
Towards Variational Flow Matching on General Geometries
-
TPP-LLM: Modeling Temporal Point Processes by Efficiently Fine-Tuning Large Language Models
-
Training Consistency Models with Variational Noise Coupling
-
Trustworthy Image Super-Resolution via Generative Pseudoinverse
-
Unifying Autoregressive And Diffusion-Based Sequence Generation
-
Unifying Causal and Object-centric Representation Learning allows Causal Composition
-
UniMoT: Unified Molecule-Text Language Model with Discrete Token Representation
-
Unpaired Point Cloud Completion using Unbalanced Optimal Transport Map
-
Variational Rectified Flow Matching
-
Video Latent Flow Matching: Optimal Polynomial Projections for Video Interpolation and Extrapolation
-
Weak-to-Strong Diffusion with Reflection
-
Your Image is Secretly the Last Frame of a Pseudo Video