SCOREBED isolates EIG double intractability in a policy-independent score-matching stage, then trains design policies with a singly intractable gradient estimator, enabling cheap multi-policy selection.
arXiv preprint arXiv:2512.22999 , year=
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
abstract
We consider problems of parameter estimation where design variables can be actively optimized to maximize information gain. To this end, we introduce JADAI, a framework that jointly amortizes Bayesian adaptive design and inference by training a policy, a history network, and an inference network end-to-end. The networks minimize a generic loss that aggregates incremental reductions in posterior error along experimental sequences. Inference networks are instantiated with diffusion-based posterior estimators that can approximate high-dimensional and multimodal posteriors at every experimental step. Across standard adaptive design benchmarks, JADAI achieves superior or competitive performance.
years
2026 5representative citing papers
Action-BED recasts BED as expected future loss on actions, producing singly intractable objectives jointly optimized for design and action policies via stochastic gradients without explicit posterior estimation.
FairBED quantifies dataset fairness as uninformative about sensitive attributes and uses fairness-aware BED to gather data yielding better fairness-accuracy trade-offs than random or standard BED acquisition.
POLAR uses pretrained predictive foundation models as fixed belief-state encoders and trains only a lightweight policy head on top for amortised Bayesian experimental design, optimisation, and active learning.
Embedding selection mechanisms into generative simulators enables amortized Bayesian inference to produce debiased, well-calibrated posteriors without tractable likelihoods.
citing papers explorer
-
Bayesian Experimental Design via Score Matching
SCOREBED isolates EIG double intractability in a policy-independent score-matching stage, then trains design policies with a singly intractable gradient estimator, enabling cheap multi-policy selection.
-
Action-BED: Task-Driven Bayesian Experimental Design with Singly Intractable Objectives
Action-BED recasts BED as expected future loss on actions, producing singly intractable objectives jointly optimized for design and action policies via stochastic gradients without explicit posterior estimation.
-
FairBED: A Bayesian Experimental Design Approach to Gathering Fairer Data
FairBED quantifies dataset fairness as uninformative about sensitive attributes and uses fairness-aware BED to gather data yielding better fairness-accuracy trade-offs than random or standard BED acquisition.
-
Efficient Adaptive Data Acquisition via Pretrained Belief Representations
POLAR uses pretrained predictive foundation models as fixed belief-state encoders and trains only a lightweight policy head on top for amortised Bayesian experimental design, optimisation, and active learning.
-
Overcoming Selection Bias in Statistical Studies With Amortized Bayesian Inference
Embedding selection mechanisms into generative simulators enables amortized Bayesian inference to produce debiased, well-calibrated posteriors without tractable likelihoods.