mlsanityData to 14 Oct 2026

SPREAD: Sampling-based Pareto front Refinement via Efficient Adaptive Diffusion

arXiv 2509.21058 ↗First posted 25 Sep 2025cs.LGcited by 2

On the map: Diffusion models → Diffusion foundations and sampling

Abstract (arXiv)

Developing efficient multi-objective optimization methods to compute the Pareto set of optimal compromises between conflicting objectives remains a key challenge, especially for large-scale and expensive problems. To bridge this gap, we introduce SPREAD, a generative framework based on Denoising Diffusion Probabilistic Models (DDPMs). SPREAD first learns a conditional diffusion process over points sampled from the decision space and then, at each reverse diffusion step, refines candidates via a sampling scheme that uses an adaptive multiple gradient descent-inspired update for fast convergence alongside a Gaussian RBF-based repulsion term for diversity. Empirical results on multi-objective optimization benchmarks, including offline and Bayesian surrogate-based settings, show that SPREAD matches or exceeds leading baselines in efficiency, scalability, and Pareto front coverage. Code is available at https://github.com/safe-autonomous-systems/moo-spread .

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Placed in Diffusion foundations and sampling, one of 14 directions in Diffusion models. Builds on 38 of its 69 references that are in the map. First posted on arXiv 25 Sep 2025.

Builds on

38 of 69 references.

Closest ancestors · 8

Same direction, newest first

  1. 2025Preference-Guided Diffusion for Multi-Objective Offline Optimization21 Mar
  2. 2024Diffusion Model for Data-Driven Black-Box Optimization20 Mar
  3. 2023Diffusion Models for Black-Box Optimization12 Jun
  4. 2022Scalable Diffusion Models with Transformers19 Dec
  5. 2022Classifier-Free Diffusion Guidance26 Jul
  6. 2021Diffusion Models Beat GANs on Image Synthesis11 May
  7. 2021Improved Denoising Diffusion Probabilistic Models18 Feb
  8. 2020Denoising Diffusion Probabilistic Models19 Jun

Other references · 30

  1. 2020DiffWave: A Versatile Diffusion Model for Audio Synthesiscited by 1,354
  2. 2020Gradient Surgery for Multi-Task Learningcited by 1,256
  3. 2017GradNorm: Gradient Normalization for Adaptive Loss Balancing in Deep Multitask Networkscited by 834
  4. 2018Multi-Task Learning as Multi-Objective Optimizationcited by 821
  5. 2022Equivariant Diffusion for Molecule Generation in 3Dcited by 677
  6. 2022Diffsound: Discrete Diffusion Model for Text-to-sound Generationcited by 257
  7. 2020pymoo: Multi-objective Optimization in Pythoncited by 200
  8. 2018A Flexible Framework for Multi-Objective Bayesian Optimization using Random Scalarizationscited by 99
  9. 2021Conservative Objective Models for Effective Offline Model-Based Optimizationcited by 88
  10. 2021Multi-Objective Bayesian Optimization over High-Dimensional Search Spacescited by 87
  11. 2020An Easy-to-use Real-world Multi-objective Optimization Problem Suitecited by 71
  12. 2024Smooth Tchebycheff Scalarization for Multi-Objective Optimizationcited by 54
18 more
  1. 2022Pareto Set Learning for Expensive Multi-Objective Optimizationcited by 51
  2. 2021RoMA: Robust Model Adaptation for Offline Model-based Optimizationcited by 39
  3. 2022Data-Driven Offline Decision-Making via Invariant Representation Learningcited by 31
  4. 2023Importance-aware Co-teaching for Offline Model-based Optimizationcited by 28
  5. 2023Parallel-mentoring for Offline Model-based Optimizationcited by 26
  6. 2024ParetoFlow: Guided Flows in Multi-Objective Optimizationcited by 17
  7. 2021Multi-Objective Learning to Predict Pareto Fronts Using Hypervolume Maximizationcited by 16
  8. 2024Offline Multi-Objective Optimizationcited by 14
  9. 2024Pareto Front-Diverse Batch Multi-Objective Bayesian Optimizationcited by 13
  10. 2016Gradient-Based Multiobjective Optimization with Uncertaintiescited by 11
  11. 2023qPOTS: Efficient batch multiobjective Bayesian optimization via Pareto optimal Thompson samplingcited by 11
  12. 2024PMGDA: A Preference-based Multiple Gradient Descent Algorithmcited by 10
  13. 2024Expensive Multi-Objective Bayesian Optimization Based on Diffusion Modelscited by 10
  14. 2024LibMOON: A Gradient-based MultiObjective OptimizatioN Library in PyTorchcited by 10
  15. 2024Multi-objective Deep Learning: Taxonomy and Survey of the State of the Artcited by 6
  16. 2024Improving Pareto Set Learning for Expensive Multi-objective Optimization via Stein Variational Hypernetworkscited by 4
  17. 2021Derivative-Free Multiobjective Trust Region Descent Method Using Radial Basis Function Surrogate Modelscited by 3
  18. 2025Multi-objective methods in Federated Learning: A survey and taxonomycited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026Large Language Models as Amortized Pareto-Front Generators for Constrained Bi-Objective Convex Optimizationcited by 0
  2. 2026ParetoTransport: Generative Optimization by Mass Transport Toward The Pareto Frontcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Pareto-Conditioned Diffusion Models for Offline Multi-Objective Optimization
  2. 2026ParetoPilot: Zero-Surrogate Offline Multi-Objective Optimization via Infer-Perturb-Guide Diffusion
  3. 2026Learning Where to Steer: Noise-Space Geometry for Efficient Offline Multi-Objective Optimization with Generative Models
  4. 2026Global Convergence of Sampling-Based Nonconvex Optimization through Diffusion-Style Smoothing
  5. 2025Test-time Alignment of Diffusion Models without Reward Over-optimization
  6. 2025A Denoising Diffusion-Based Evolutionary Algorithm Framework: Application to the Maximum Independent Set Problem
  7. 2026Multi-Objective Learning for Diffusion Models: A Statistical Theory under Semi-Supervised Learning
  8. 2026Learn Feasibility Once, Optimize All Objectives: Derivative-Free Diffusion Models for Chance-Constrained Programming