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

Training Neural Samplers with Reverse Diffusive KL Divergence

About

Training generative models to sample from unnormalized density functions is an important and challenging task in machine learning. Traditional training methods often rely on the reverse Kullback-Leibler (KL) divergence due to its tractability. However, the mode-seeking behavior of reverse KL hinders effective approximation of multi-modal target distributions. To address this, we propose to minimize the reverse KL along diffusion trajectories of both model and target densities. We refer to this objective as the reverse diffusive KL divergence, which allows the model to capture multiple modes. Leveraging this objective, we train neural samplers that can efficiently generate samples from the target distribution in one step. We demonstrate that our method enhances sampling performance across various Boltzmann distributions, including both synthetic multi-modal densities and n-body particle systems.

Jiajun He, Wenlin Chen, Mingtian Zhang, David Barber, Jos\'e Miguel Hern\'andez-Lobato• 2024

Related benchmarks

TaskDatasetResultRank
n-body particle system samplingDW-4 d = 8
W2 Distance1.566
20
n-body particle system samplingLJ-13 (d = 39)
W2 Distance4.058
13
n-body particle system samplingLJ-55 d = 165
W218.05
10
Sampling n-Body Particle SystemsLJ-55
Time per Step (ms)368.4
8
Showing 4 of 4 rows

Other info

Follow for update