UPS and electric maintenance is scheduled for Wednesday, November 27th, 2024, 08:30 - 12:00. A downtime of this service might occur for up to 30 minutes.

21–25 Oct 2024
Bethe Center
Europe/Berlin timezone

Importance weights distribution in Neural Samplers

23 Oct 2024, 11:10
40m
Wegelerstr. 10 - Seminar Room 2.019 - 53115 Bonn (Bethe Center)

Wegelerstr. 10 - Seminar Room 2.019 - 53115 Bonn

Bethe Center

Speaker

Prof. Piotr Bialas

Description

Neural samplers in general generate configurations from distribution that only approximates the desired target distribution. The importance weights, which are the quotients of the target probability of the sample to its actual probability, do account for this discrepancy and permit the correction of the distribution either by reweighting the samples or accepting/rejecting them using the Metropolis algorithm. Ideally, we would like those weights to be distributed around one with a small variance. It turns out however that, in the case of poorly trained sampler, this distribution can be long tailed, even to the point of having infinite variance. In my talk I will discuss possible causes of this behavior and its connection with mode collapse.

Presentation materials