Pith. sign in

REVIEW 3 cited by

Sequential Neural Score Estimation: Likelihood-Free Inference with Conditional Score Based Diffusion Models

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2210.04872 v3 pith:622WPHKK submitted 2022-10-10 stat.ML cs.LG

Sequential Neural Score Estimation: Likelihood-Free Inference with Conditional Score Based Diffusion Models

classification stat.ML cs.LG
keywords posteriorsequentialscoreestimationmethodmodelsneuralscore-based
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

We introduce Sequential Neural Posterior Score Estimation (SNPSE), a score-based method for Bayesian inference in simulator-based models. Our method, inspired by the remarkable success of score-based methods in generative modelling, leverages conditional score-based diffusion models to generate samples from the posterior distribution of interest. The model is trained using an objective function which directly estimates the score of the posterior. We embed the model into a sequential training procedure, which guides simulations using the current approximation of the posterior at the observation of interest, thereby reducing the simulation cost. We also introduce several alternative sequential approaches, and discuss their relative merits. We then validate our method, as well as its amortised, non-sequential, variant on several numerical examples, demonstrating comparable or superior performance to existing state-of-the-art methods such as Sequential Neural Posterior Estimation (SNPE).

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 3 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Amortized Simulation-Based Inference in Generalized Bayes via Neural Posterior Estimation

    stat.ML 2026-01 unverdicted novelty 7.0

    Introduces the first amortized neural posterior estimator conditioned on both data and temperature β for generalized Bayesian inference, matching MCMC performance on standard SBI benchmarks.

  2. Flow-based generative models for amortized Bayesian inference in regression and inverse PDE problems

    physics.comp-ph 2026-06 unverdicted novelty 6.0

    Flow-ABI trains flow-matching models on historical data to produce a set-conditioned functional posterior sampler that delivers near-real-time Bayesian inference for regression and inverse PDE tasks without per-observ...

  3. Overcoming Selection Bias in Statistical Studies With Amortized Bayesian Inference

    stat.ML 2026-04 unverdicted novelty 6.0

    Embedding selection mechanisms into generative simulators enables amortized Bayesian inference to produce debiased, well-calibrated posteriors without tractable likelihoods.