Research map / Probabilistic and scientific modeling / Variational and generative sampling

High-Dimensional Bayesian Optimisation with Gaussian Process Prior Variational Autoencoders

ICLR 2025 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Structured latent VAEs

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes combining Bayesian optimisation with a GP-prior VAE (conditional generative model) to perform high-dimensional BO in a latent space.

Problem

Bayesian optimisation with GP surrogates scales poorly to high-dimensional inputs; prior VAE-based latent BO methods project data to a low-dimensional latent space.

Approach

Uses a conditional generative model with a GP prior in the VAE latent space (GP prior VAE) so that the latent representation is structured for the downstream GP surrogate, enabling BO in the inferred latent space.

Experimental setup

Builds on: variational autoencoder latent space bayesian optimisation

Closest papers by meaning

• Bayesian Basis Function Approximation for Scalable Gaussian Process Priors in Deep Generative Models (ICML 2025 · similarity 0.91)
• PG-LBO: Enhancing High-Dimensional Bayesian Optimization with Pseudo-Label and Gaussian Process Guidance (AAAI 2024 · similarity 0.91)
• Neighbour-Driven Gaussian Process Variational Autoencoders for Scalable Structured Latent Modelling (ICML 2025 · similarity 0.90)
• Bayesian Optimization via Continual Variational Last Layer Training (ICLR 2025 · similarity 0.88)
• Standard Gaussian Process is All You Need for High-Dimensional Bayesian Optimization (ICLR 2025 · similarity 0.88)
• Bayesian Optimization through Gaussian Cox Process Models for Spatio-temporal Data (ICLR 2024 · similarity 0.88)