A new ray-tracing MCMC sampler keeps ray speed constant, making it far more robust to stochastic gradients and able to sample billion-parameter neural networks on one GPU.
Gaussian Processes to speed up MCMC with automatic exploratory-exploitation effect
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
We present a two-stage Metropolis-Hastings algorithm for sampling probabilistic models, whose log-likelihood is computationally expensive to evaluate, by using a surrogate Gaussian Process (GP) model. The key feature of the approach, and the difference w.r.t. previous works, is the ability to learn the target distribution from scratch (while sampling), and so without the need of pre-training the GP. This is fundamental for automatic and inference in Probabilistic Programming Languages In particular, we present an alternative first stage acceptance scheme by marginalising out the GP distributed function, which makes the acceptance ratio explicitly dependent on the variance of the GP. This approach is extended to Metropolis-Adjusted Langevin algorithm (MALA).
fields
astro-ph.IM 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
The Ray Tracing Sampler: Bayesian Sampling of Neural Networks for Everyone
A new ray-tracing MCMC sampler keeps ray speed constant, making it far more robust to stochastic gradients and able to sample billion-parameter neural networks on one GPU.