mlsanityData to 14 Oct 2026

Mitigating mode collapse in normalizing flows by annealing with an adaptive schedule: Application to parameter estimation

arXiv 2505.03652 ↗First posted 6 May 2025cs.LG physics.comp-ph physics.data-ancited by 6

Not on the map. Why?

Abstract (arXiv)

Normalizing flows (NFs) provide uncorrelated samples from complex distributions, making them an appealing tool for parameter estimation. However, the practical utility of NFs remains limited by their tendency to collapse to a single mode of a multimodal distribution. In this study, we show that annealing with an adaptive schedule based on the effective sample size (ESS) can mitigate mode collapse. We demonstrate that our approach can converge the marginal likelihood for a biochemical oscillator model fit to time-series data in ten-fold less computation time than a widely used ensemble Markov chain Monte Carlo (MCMC) method. We show that the ESS can also be used to reduce variance by pruning the samples. We expect these developments to be of general use for sampling with NFs and discuss potential opportunities for further improvements.

Key facts. Cited by 6 papers in the map (data through 14 Oct 2026). Builds on 19 of its 61 references that are in the map. First posted on arXiv 6 May 2025.

Builds on

19 of 61 references.

References · 19

  1. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  2. 2016Density estimation using Real NVPcited by 2,042
  3. 2018Glow: Generative Flow with Invertible 1x1 Convolutionscited by 1,545
  4. 2016Variational Inference: A Review for Statisticianscited by 1,472
  5. 2019Normalizing Flows for Probabilistic Modeling and Inferencecited by 1,323
  6. 2023Stochastic Interpolants: A Unifying Framework for Flows and Diffusionscited by 939
  7. 2017Masked Autoregressive Flow for Density Estimationcited by 684
  8. 2019Neural Spline Flowscited by 596
  9. 2018FFJORD: Free-form Continuous Dynamics for Scalable Reversible Generative Modelscited by 594
  10. 2022Flow Annealed Importance Sampling Bootstrapcited by 127
  11. 2018Solving Statistical Mechanics Using Variational Autoregressive Networkscited by 100
  12. 2021Stiff Neural Ordinary Differential Equationscited by 78
7 more
  1. 2023Learning Interpolations between Boltzmann Densitiescited by 43
  2. 2015Pareto Smoothed Importance Samplingcited by 42
  3. 2023Detecting and Mitigating Mode-Collapse for Flow-based Sampling of Lattice Field Theoriescited by 25
  4. 2022Gradients should stay on Path: Better Estimators of the Reverse- and Forward KL Divergence for Normalizing Flowscited by 23
  5. 2021Efficient Bayesian Sampling Using Normalizing Flows to Assist Markov Chain Monte Carlo Methodscited by 16
  6. 2022Deterministic Langevin Monte Carlo with Normalizing Flows for Bayesian Inferencecited by 10
  7. 2023Fixed-kinetic Neural Hamiltonian Flows for enhanced interpretability and reduced complexitycited by 4

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 6.

6 papers.

  1. 2026Annealing in variational inference mitigates mode collapse: A theoretical study on Gaussian mixturescited by 4
  2. 2026Leveraging generative models to assist Monte Carlo samplingcited by 1
  3. 2026Jeffreys Flow: Robust Boltzmann Generators for Rare Event Sampling via Parallel Tempering Distillationcited by 0
  4. 2026Generative Nested Sampling of Atomistic Thermodynamic Landscapescited by 0
  5. 2026Machine learning kinetics from molecular dynamics datacited by 0
  6. 2026SPIBER: Reconstructing Free Energy Landscapes from Short, Unconverged Trajectories with Generative Flow Networkscited by 0