mlsanityData to 14 Oct 2026

Path-Gradient Estimators for Continuous Normalizing Flows

arXiv 2206.09016 ↗First posted 17 Jun 2022cs.LG stat.MLcited by 11

Not on the map. Why?

Abstract (arXiv)

Recent work has established a path-gradient estimator for simple variational Gaussian distributions and has argued that the path-gradient is particularly beneficial in the regime in which the variational distribution approaches the exact target distribution. In many applications, this regime can however not be reached by a simple Gaussian variational distribution. In this work, we overcome this crucial limitation by proposing a path-gradient estimator for the considerably more expressive variational family of continuous normalizing flows. We outline an efficient algorithm to calculate this estimator and establish its superior performance empirically.

Key facts. Cited by 11 papers in the map (data through 14 Oct 2026). Builds on 21 of its 30 references that are in the map. First posted on arXiv 17 Jun 2022.

Builds on

21 of 30 references.

References · 21

  1. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  2. 2013Auto-Encoding Variational Bayescited by 4,053
  3. 2018Neural Ordinary Differential Equationscited by 3,779
  4. 2016Improving Variational Inference with Inverse Autoregressive Flowcited by 640
  5. 2015Importance Weighted Autoencoderscited by 456
  6. 2020Equivariant Flows: Exact Likelihood Generative Learning for Symmetric Densitiescited by 250
  7. 2018Neural Importance Samplingcited by 155
  8. 2020Equivariant flow-based sampling for lattice gauge theorycited by 103
  9. 2018Solving Statistical Mechanics Using Variational Autoregressive Networkscited by 100
  10. 2017Sticking the Landing: Simple, Lower-Variance Gradient Estimators for Variational Inferencecited by 98
  11. 2018Sylvester Normalizing Flows for Variational Inferencecited by 72
  12. 2014Reweighted Wake-Sleepcited by 66
9 more
  1. 2018Doubly Reparameterized Gradient Estimators for Monte Carlo Objectivescited by 41
  2. 2020Advances in Black-Box VI: Normalizing Flows, Importance Weighting, and Optimizationcited by 37
  3. 2021Introduction to Normalizing Flows for Lattice Field Theorycited by 35
  4. 2021Scaling Up Machine Learning For Quantum Field Theory with Equivariant Continuous Flowscited by 24
  5. 2020On the Difficulty of Unbiased Alpha Divergence Minimizationcited by 21
  6. 2021Generalized Doubly Reparameterized Gradient Estimatorscited by 14
  7. 2021Empirical Evaluation of Biased Methods for Alpha Divergence Minimizationcited by 10
  8. 2021Machine Learning Trivializing Maps: A First Step Towards Understanding How Flow-Based Samplers Scale Upcited by 5
  9. 2021Machine Learning and Variational Algorithms for Lattice Field Theorycited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2022: 2; 2023: 4; 2024: 2; 2025: 2; 2026: 1.

11 papers.

  1. 2023Diffusion Generative Flow Samplers: Improving learning signals through partial trajectory optimizationcited by 57
  2. 2022Learning Lattice Quantum Field Theories with Equivariant Continuous Flowscited by 37
  3. 2022Gradients should stay on Path: Better Estimators of the Reverse- and Forward KL Divergence for Normalizing Flowscited by 23
  4. 2023Advances in machine-learning-based sampling motivated by lattice quantum chromodynamicscited by 23
  5. 2023Sampling the lattice Nambu-Goto string using Continuous Normalizing Flowscited by 12
  6. 2023Bridging the Gap Between Variational Inference and Wasserstein Gradient Flowscited by 8
  7. 2024Fast and Unified Path Gradient Estimators for Normalizing Flowscited by 5
  8. 2025Path Gradients after Flow Matchingcited by 3
  9. 2025SESaMo: Symmetry-Enforcing Stochastic Modulation for Normalizing Flowscited by 2
  10. 2024Multi-Lattice Sampling of Quantum Field Theories via Neural Operator-based Flowscited by 1
  11. 2026Scaling Online Optimal Transport Minibatches for Flow Matchingcited by 0