SlotFlow: Amortized Trans-Dimensional Inference with Slot-Based Normalizing Flows

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Houba, Niklas, Giarda, Giovanni, Speri, Lorenzo
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911292260876288
author Houba, Niklas
Giarda, Giovanni
Speri, Lorenzo
author_facet Houba, Niklas
Giarda, Giovanni
Speri, Lorenzo
contents Inferring the number of distinct components contributing to an observation, while simultaneously estimating their parameters, remains a long-standing challenge across signal processing, astrophysics, and neuroscience. Classical trans-dimensional Bayesian methods such as Reversible Jump Markov Chain Monte Carlo (RJMCMC) provide asymptotically exact inference but can be computationally expensive. Instead, modern deep learning provides a faster alternative to inference but typically assume fixed component counts, sidestepping the core challenge of trans-dimensionality. To address this, we introduce SlotFlow, a deep learning architecture for trans-dimensional amortized inference. The architecture processes time-series observations, which we represent jointly in the frequency and time domains through parallel encoders. A classifier produces a distribution over component counts K, and its MAP estimate specifies the number of slots instantiated. Each slot is parameterized by a shared conditional normalizing flow trained via permutation-invariant Hungarian matching. On sinusoidal decomposition with up to 10 overlapping components and Gaussian noise, SlotFlow achieves 99.85% cardinality accuracy and well-calibrated parameter posteriors, with systematic biases well below one posterior standard deviation. Direct comparison with RJMCMC shows close agreement in amplitude and phase, with Wasserstein distances $W_2 < 0.01$ and $< 0.03$, indicating that shared global context captures inter-component structure despite a factorized posterior. Frequency posteriors remain centered but exhibit 2-3x broader intervals, consistent with an encoder bottleneck in retaining long-baseline phase coherence. The method delivers a $\sim 10^6\times$ speedup over RJMCMC, suggesting applicability to time-critical workflows in gravitational-wave astronomy, neural spike sorting, and object-centric vision.
format Preprint
id arxiv_https___arxiv_org_abs_2511_23228
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SlotFlow: Amortized Trans-Dimensional Inference with Slot-Based Normalizing Flows
Houba, Niklas
Giarda, Giovanni
Speri, Lorenzo
Instrumentation and Methods for Astrophysics
General Relativity and Quantum Cosmology
Data Analysis, Statistics and Probability
Inferring the number of distinct components contributing to an observation, while simultaneously estimating their parameters, remains a long-standing challenge across signal processing, astrophysics, and neuroscience. Classical trans-dimensional Bayesian methods such as Reversible Jump Markov Chain Monte Carlo (RJMCMC) provide asymptotically exact inference but can be computationally expensive. Instead, modern deep learning provides a faster alternative to inference but typically assume fixed component counts, sidestepping the core challenge of trans-dimensionality. To address this, we introduce SlotFlow, a deep learning architecture for trans-dimensional amortized inference. The architecture processes time-series observations, which we represent jointly in the frequency and time domains through parallel encoders. A classifier produces a distribution over component counts K, and its MAP estimate specifies the number of slots instantiated. Each slot is parameterized by a shared conditional normalizing flow trained via permutation-invariant Hungarian matching. On sinusoidal decomposition with up to 10 overlapping components and Gaussian noise, SlotFlow achieves 99.85% cardinality accuracy and well-calibrated parameter posteriors, with systematic biases well below one posterior standard deviation. Direct comparison with RJMCMC shows close agreement in amplitude and phase, with Wasserstein distances $W_2 < 0.01$ and $< 0.03$, indicating that shared global context captures inter-component structure despite a factorized posterior. Frequency posteriors remain centered but exhibit 2-3x broader intervals, consistent with an encoder bottleneck in retaining long-baseline phase coherence. The method delivers a $\sim 10^6\times$ speedup over RJMCMC, suggesting applicability to time-critical workflows in gravitational-wave astronomy, neural spike sorting, and object-centric vision.
title SlotFlow: Amortized Trans-Dimensional Inference with Slot-Based Normalizing Flows
topic Instrumentation and Methods for Astrophysics
General Relativity and Quantum Cosmology
Data Analysis, Statistics and Probability
url https://arxiv.org/abs/2511.23228