Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Path Integral Sampler: a stochastic control approach for sampling

About

We present Path Integral Sampler~(PIS), a novel algorithm to draw samples from unnormalized probability density functions. The PIS is built on the Schr\"odinger bridge problem which aims to recover the most likely evolution of a diffusion process given its initial distribution and terminal distribution. The PIS draws samples from the initial distribution and then propagates the samples through the Schr\"odinger bridge to reach the terminal distribution. Applying the Girsanov theorem, with a simple prior diffusion, we formulate the PIS as a stochastic optimal control problem whose running cost is the control energy and terminal cost is chosen according to the target distribution. By modeling the control as a neural network, we establish a sampling algorithm that can be trained end-to-end. We provide theoretical justification of the sampling quality of PIS in terms of Wasserstein distance when sub-optimal control is used. Moreover, the path integrals theory is used to compute importance weights of the samples to compensate for the bias induced by the sub-optimality of the controller and time-discretization. We experimentally demonstrate the advantages of PIS compared with other start-of-the-art sampling methods on a variety of tasks.

Qinsheng Zhang, Yongxin Chen• 2021

Related benchmarks

TaskDatasetResultRank
Unconditional modeling25GMM d = 2
Delta Log Z1.799
30
Unconditional modelingFunnel d = 10
Delta log Z0.534
30
Unconditional modelingManywell d = 32
Δ log Z13.19
29
n-body particle system samplingDW-4 d = 8
W2 Distance0.68
29
Target Distribution SamplingFunnel 10D
Sinkhorn Distance155.2
29
n-body particle system samplingLJ-13 (d = 39)
W2 Distance1.93
21
Toy target distribution samplingGMM40 d = 50
W2 (Entropy Regulated, eps=0.05)1.04e+4
18
n-body particle system samplingLJ-55 d = 165
W24.79
16
Conditional SamplingMNIST pretrained VAE decoder (test)
log Z-99.89
15
Unconditional modelingLog-Gaussian Cox process d = 1600
Delta log Z471.4
13
Showing 10 of 30 rows

Other info

Follow for update