mlsanityData to 7 Oct 2026

Latent Space Refinement for Deep Generative Models

arXiv 2106.00792 ↗First posted 1 Jun 2021stat.ML cs.LG hep-excited by 10

Not on the map. Why?

Abstract (arXiv)

Deep generative models are becoming widely used across science and industry for a variety of purposes. A common challenge is achieving a precise implicit or explicit representation of the data probability density. Recent proposals have suggested using classifier weights to refine the learned density of deep generative models. We extend this idea to all types of generative models and show how latent space refinement via iterated generative modeling can circumvent topological obstructions and improve precision. This methodology also applies to cases were the target model is non-differentiable and has many internal latent dimensions which must be marginalized over before refinement. We demonstrate our Latent Space Refinement (LaSeR) protocol on a variety of examples, focusing on the combinations of Normalizing Flows and Generative Adversarial Networks.

Key facts. Cited by 10 papers in the map (data through 7 Oct 2026). Builds on 42 of its 97 references that are in the map. First posted on arXiv 1 Jun 2021.

Builds on

42 of 97 references.

References · 42

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  3. 2015Variational Inference with Normalizing Flowscited by 2,161
  4. 2016Density estimation using Real NVPcited by 2,029
  5. 2019An Introduction to Variational Autoencoderscited by 997
  6. 2017Generative Adversarial Networks: An Overviewcited by 588
  7. 2015Generative Moment Matching Networkscited by 252
  8. 2019Diagnosing and Enhancing VAE Modelscited by 192
  9. 2019Flow-based generative models for Markov chain Monte Carlo in lattice field theorycited by 147
  10. 2020Flows for simultaneous manifold learning and density estimationcited by 117
  11. 2020Equivariant flow-based sampling for lattice gauge theorycited by 102
  12. 2018Iterative Amortized Inferencecited by 93
30 more
  1. 2018Semi-Amortized Variational Autoencoderscited by 83
  2. 2017CaloGAN: Simulating 3D High Energy Particle Showers in Multi-Layer Electromagnetic Calorimeters with Generative Adversarial Networkscited by 78
  3. 2020Anomaly Detection with Density Estimationcited by 71
  4. 2020Your GAN is Secretly an Energy-based Model and You Should use Discriminator Driven Latent Samplingcited by 71
  5. 2020Augmented Normalizing Flows: Bridging the Gap Between Generative Flows and Latent Variable Modelscited by 66
  6. 2017Accelerating Science with Generative Adversarial Networks: An Application to 3D Particle Showers in Multi-Layer Calorimeterscited by 55
  7. 2017Learning Particle Physics by Example: Location-Aware Generative Adversarial Networks for Physics Synthesiscited by 46
  8. 2019Calorimetry with Deep Learning: Particle Simulation and Reconstruction for Collider Physicscited by 42
  9. 2020Variational Autoencoders for Anomalous Jet Taggingcited by 33
  10. 2019OmniFold: A Method to Simultaneously Unfold All Observablescited by 31
  11. 2019How to GAN away Detector Effectscited by 28
  12. 2019LHC analysis-specific datasets with Generative Adversarial Networkscited by 26
  13. 2020i-flow: High-dimensional Integration and Sampling with Normalizing Flowscited by 26
  14. 2018Generative Models for Fast Calorimeter Simulation.LHCb casecited by 25
  15. 2017Controlling Physical Attributes in GAN-Accelerated Simulation of Electromagnetic Calorimeterscited by 23
  16. 2020DCTRGAN: Improving the Precision of Generative Models with Reweightingcited by 22
  17. 2019Neural Networks for Full Phase-space Reweighting and Parameter Tuningcited by 20
  18. 2021Bump Hunting in Latent Spacecited by 19
  19. 2020End-to-end Sinkhorn Autoencoder with Noise Generatorcited by 17
  20. 2021Understanding Event-Generation Networks via Uncertaintiescited by 17
  21. 2021E Pluribus Unum Ex Machina: Learning from Many Collider Events at Oncecited by 15
  22. 2020Variational Autoencoders for Jet Simulationcited by 14
  23. 2020Probabilistic Autoencodercited by 12
  24. 2019Lund jet images from generative and cycle-consistent adversarial networkscited by 11
  25. 2020Graph Generative Adversarial Networks for Sparse Data Generation in High Energy Physicscited by 11
  26. 2019DeepRICH: Learning Deeply Cherenkov Detectorscited by 9
  27. 2020Simulating the Time Projection Chamber responses at the MPD detector using Generative Adversarial Networkscited by 8
  28. 2018Image-based model parameter optimization using Model-Assisted Generative Adversarial Networkscited by 7
  29. 2018GANs for generating EFT modelscited by 6
  30. 2021Reduced Precision Strategies for Deep Learning: A High Energy Physics Generative Adversarial Network Use Casecited by 6

Built on by

Cited by year

Citing papers per year: 2021: 2; 2022: 1; 2023: 5; 2024: 2.

10 papers.

  1. 2021CaloFlow: Fast and Accurate Generation of Calorimeter Showers with Normalizing Flowscited by 46
  2. 2021Generative Networks for Precision Enthusiastscited by 21
  3. 2023Deep Generative Models for Detector Signature Simulation: A Taxonomic Reviewcited by 18
  4. 2022New directions for surrogate models and differentiable programming for High Energy Physics detector simulationcited by 16
  5. 2023High-dimensional and Permutation Invariant Anomaly Detectioncited by 12
  6. 2023Calorimeter shower superresolutioncited by 7
  7. 2023ELSA -- Enhanced latent spaces for improved collider simulationscited by 5
  8. 2024Calibrating Bayesian Generative Machine Learning for Bayesiamplificationcited by 5
  9. 2023Steering Language Generation: Harnessing Contrastive Expert Guidance and Negative Prompting for Coherent and Diverse Synthetic Data Generationcited by 2
  10. 2024Extrapolating Jet Radiation with Autoregressive Transformerscited by 2