REVIEW 20 cited by
An Overview of Diffusion Models: Applications, Guided Generation, Statistical Rates and Optimization
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
An Overview of Diffusion Models: Applications, Guided Generation, Statistical Rates and Optimization
read the original abstract
Diffusion models, a powerful and universal generative AI technology, have achieved tremendous success in computer vision, audio, reinforcement learning, and computational biology. In these applications, diffusion models provide flexible high-dimensional data modeling, and act as a sampler for generating new samples under active guidance towards task-desired properties. Despite the significant empirical success, theory of diffusion models is very limited, potentially slowing down principled methodological innovations for further harnessing and improving diffusion models. In this paper, we review emerging applications of diffusion models, understanding their sample generation under various controls. Next, we overview the existing theories of diffusion models, covering their statistical properties and sampling capabilities. We adopt a progressive routine, beginning with unconditional diffusion models and connecting to conditional counterparts. Further, we review a new avenue in high-dimensional structured optimization through conditional diffusion models, where searching for solutions is reformulated as a conditional sampling problem and solved by diffusion models. Lastly, we discuss future directions about diffusion models. The purpose of this paper is to provide a well-rounded theoretical exposure for stimulating forward-looking theories and methods of diffusion models.
Forward citations
Cited by 20 Pith papers
-
D$^3$-Subsidy: Online and Sequential Driver Subsidy Decision-Making for Large-Scale Ride-Hailing Market
D³-Subsidy uses prefix-conditioned diffusion to sample future trajectories from history, decodes them into city-level controls, and applies Lagrangian mapping to enforce subsidy caps, reporting gains in rides and GMV ...
-
D$^3$-Subsidy: Online and Sequential Driver Subsidy Decision-Making for Large-Scale Ride-Hailing Market
D³-Subsidy uses prefix-conditioned diffusion models to sample future trajectories and a Lagrangian mapping to produce cap-compliant city-level subsidy controls that improve rides and GMV in large ride-hailing systems.
-
Generative Diffusion Prior Distillation for Long-Context Knowledge Transfer
GDPD treats partial student features as degraded observations and uses a learned diffusion prior over teacher features to sample restorative long-context targets for improved partial time-series classification.
-
Fast Score-Based Sampling via Log-Concave Reductions
Score-based sampling reduces to a short sequence of strongly log-concave sampling problems, giving √d polylog(1/ε) complexity bounds and logarithmic dependence on the condition number for log-concave targets.
-
Improved Sample Complexity For Diffusion Model Training Without Empirical Risk Minimizer Access
The paper establishes an O(ε^{-4}) sample complexity bound for score estimation in diffusion models without requiring access to the empirical risk minimizer.
-
From Score Learning to Discretized Sampling: An End-to-End Generalization Analysis of Diffusion Models
An end-to-end TV bound for score-based diffusion models that decomposes generative error into forward truncation, reverse discretization, finite-sample generalization, and optimization gap.
-
From Similarity to Feasibility: Diffusion-Refined Retrieval-Augmented Generation for Distribution Network Optimization
A retrieval-and-diffusion warm-start pipeline cuts solve times for distribution-network optimization on most tested benchmarks while keeping solution quality near-optimal, but it is slower than direct solving on one c...
-
Nonlocal Onsager Operators and Entropy Dissipation for Finite-State Schr\"odinger Bridges
Develops nonlocal Onsager operators for terminal marginal evolution in finite-state Schrödinger bridges as gradient flows of relative entropy, with global well-posedness and convergence proved under positivity assumptions.
-
Do Less, Achieve More: Do We Need Every-Step Optimization for RL Fine-tuning of Diffusion Models?
AdaScope adaptively selects optimal RL intervention points during diffusion denoising by monitoring structural and semantic changes, delivering 66% higher performance at 59% lower cost than full-trajectory RL baselines.
-
Slowly Annealed Langevin Dynamics: Theory and Applications to Training-Free Guided Generation
Slowly Annealed Langevin Dynamics provides non-asymptotic KL-based convergence guarantees for tracking moving targets and enables training-free guided generation via a velocity-aware correction that accounts for pretr...
-
On the Robustness of Distribution Support under Diffusion Guidance
Guided diffusion generates samples near the target distribution support under exact score access, explaining its empirical success in producing plausible outputs.
-
Conditional Diffusion Guidance under Hard Constraint: A Stochastic Analysis Approach
By adding drift g(t)^2 ∇log h(t,y) with h estimated via martingale and covariation losses, diffusion samples can be hard-conditioned on an event.
-
MMaDA: Multimodal Large Diffusion Language Models
MMaDA is a unified multimodal diffusion model using mixed chain-of-thought fine-tuning and a new UniGRPO reinforcement learning algorithm that outperforms specialized models in reasoning, understanding, and text-to-im...
-
Characterizing higher-order representations through generative diffusion models explains human decoded neurofeedback performance
NERD uses RL-trained diffusion models on fMRI data to model higher-order uncertainty representations, outperforming controls and linking individual differences to neurofeedback success.
-
D$^3$-Subsidy: Online and Sequential Driver Subsidy Decision-Making for Large-Scale Ride-Hailing Market
D³-Subsidy is a prefix-conditioned diffusion model plus Lagrangian mapping that generates city-level subsidy plans from historical data and maps them to per-order incentives while respecting rate caps.
-
D$^3$-Subsidy: Online and Sequential Driver Subsidy Decision-Making for Large-Scale Ride-Hailing Market
D³-Subsidy uses prefix-conditioned diffusion models to generate future subsidy trajectories from history, decoded via inverse modules and mapped with Lagrangian duality to city-level incentives.
-
SAS: Semantic-aware Sampling for Generative Dataset Distillation
SAS adds semantic scoring with CLIP and a two-stage filter-then-diversity selection process to make generative dataset distillation produce more class-discriminative and diverse compact datasets.
-
Conditional diffusion denoising probabilistic model for super-resolution of atmospheric boundary layer large eddy simulation
A conditional diffusion model super-resolves coarse ABL LES data, recovering fine turbulent structures and Reynolds stresses accurately inside the training distribution but producing noise and over-predicted stresses ...
-
On the Robustness of Distribution Support under Diffusion Guidance
Establishes robustness of distribution support for guided diffusion processes under exact score access across DDIM, DDPM, and exponential integrator discretizations.
-
Inverse Design of Inorganic Compounds with Generative AI
A review of generative AI for inverse design of inorganic compounds, analyzing adaptations for their complexity in composition, geometry, symmetry, and electronic structure, with discussion of future benchmarks and sy...
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.