AI Paper Atlas

Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes

Distribution Transformers: Fast Approximate Bayesian Inference With On-The-Fly Prior Adaptation

George Whittle, Juliusz Ziomek, Jacob Henry Rawling, Michael A Osborne

ICML 2026 · Probabilistic and scientific modeling › Uncertainty and Gaussian processes · approach: Simulation-based inference · 0 citations (OpenAlex)

Read the paper

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add prior loss term to the training objective (ablation)Distribution Transformer · Inverse-gamma (wide meta-prior), prior KLexpected prior KL-divergence: 0.3020 -> 0.0004Improved
add prior loss term to the training objective (ablation)Distribution Transformer · Inverse-gamma (wide meta-prior), posterior KLexpected posterior KL-divergence: 0.0004 -> 0.0003Improved
increase GMM components from 2 to 5 in the DTDistribution Transformer · Inverse-gamma (narrow meta-prior)expected KL-divergence: 0.0044 -> 0.0004Improved
increase GMM components from 2 to 5 in the DTDistribution Transformer · Inverse-gamma (wide meta-prior)expected KL-divergence: 0.0058 -> 0.0003Improved
test on unseen shifted meta-prior instead of training meta-prior (ablation)DT-5 · Meta-prior misspecificationexpected KL-divergence: 0.0003 -> 0.0054Hurt
halve GMM components at inference time (10 -> 5) (ablation)DT trained with 10 components · 8-dim linear Gaussianexpected posterior KL-divergence: 0.0873 -> 0.0985Hurt
double GMM components at inference time (10 -> 20) (ablation)DT trained with 10 components · 8-dim linear Gaussianexpected posterior KL-divergence: 0.0873 -> 0.0946Hurt
use 2 components at inference time instead of 10 (ablation)DT trained with 10 components · 8-dim linear Gaussianexpected posterior KL-divergence: 0.0873 -> 0.1494Hurt
replace EKF with Distribution Transformer for filteringDT · Bayesian sensor fusionexpected NLL: 95.9 -> -0.197Improved
replace particle filter with Distribution TransformerDT · Bayesian sensor fusioniteration time: 0.818 -> 0.017Improved
use DT-5 rather than a PFN with 5000 Riemann bucketsDT-5 · Inverse-gamma (wide meta-prior)expected KL-divergence: 0.2935 -> 0.0003Improved
use DT instead of ACE for GP hyperposteriorDT · GP hyperposteriorexpected NLL: 0.35 -> 0.31Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy