Towards Universal Neural Likelihood Inference
Abstract
We introduce universal neural likelihood inference (UNLI): enabling a single model to provide data-grounded, conditional likelihood predictions for arbitrary targets given any collection of observed features, across diverse domains and tasks. To achieve UNLI over heterogeneous tabular data, we develop the Arbitrary Set-based Permutation-Invariant Reasoning Engine (ASPIRE) model. Our design addresses critical gaps in existing approaches to merge semantic-understanding capabilities and generalised numerical feature reasoning within a zero-shot capable framework. Trained on over 1,400 real diverse datasets spanning various domains, ASPIRE achieves 15\% higher F1 scores and 85\% lower RMSE than existing tabular foundation models in zero-shot and few-shot settings. Lastly, this work introduces open-world active feature acquisition, where we leverage the UNLI capabilities of ASPIRE to adeptly determine next feature-values to observe to improve inference time prediction accuracies.
Cite
@article{arxiv.2508.09100,
title = {Towards Universal Neural Likelihood Inference},
author = {Shreyas Bhat Brahmavar and Yang Li and Qiyang Liu and Shashank Srivastava and Junier Oliva},
journal= {arXiv preprint arXiv:2508.09100},
year = {2026}
}