← Back to all news
arXiv Machine Learning By Vikas Kanaujia, Riyansha Singh, Shashank Sharma, Vipul Arora

Symbolic Density Estimators for Unnormalized Distributions

Read the original on arXiv Machine Learning →

The paper introduces a framework that combines deep generative models with symbolic regression to estimate symbolic expressions for unnormalized probability density functions directly from observed samples. By incorporating domain-specific priors—such as interaction ranges and a set of primitive functions—the method can factorize large distributions and remain tractable. Experiments demonstrate its effectiveness on multivariate toy distributions and on lattice models from computational physics, including the XY model and φ⁴ theory, where it produces compact symbolic approximations of the Hamiltonian across scales.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.

Related stories

arXiv Machine Learning
5d ago

Active Learning of Conditional Generative Models via the Transport Neural Tangent Kernel

The paper introduces the transport Neural Tangent Kernel (tNTK) to enable active learning for conditional generative models that produce distributional outputs. By decomposing the Wasserstein test risk into mean and shape components, the authors show that mean‑based active learning misses distributional shape, and propose tNTK as a tractable kernel that captures sensitivity of the predicted distribution to model parameters. Using tNTK as a Gaussian process covariance, they greedily select inputs that minimize posterior variance, and demonstrate that this distribution‑aware strategy outperforms mean‑based NTK baselines across several generative model types and datasets.

By Jayoung Ryu, Kyunghyun Cho, Romain Lopez
diffusionefficiency
More like this →