Parallel and Flexible Sampling from Autoregressive Models via Langevin Dynamics
This addresses the computational bottleneck in sampling from autoregressive models for applications in visual and audio domains, offering a more flexible and efficient approach.
The paper tackles the problem of slow sequential sampling from autoregressive models by proposing a parallel sampling method using Langevin dynamics on global log-likelihood, achieving competitive results in tasks like audio source separation, super-resolution, and inpainting.
This paper introduces an alternative approach to sampling from autoregressive models. Autoregressive models are typically sampled sequentially, according to the transition dynamics defined by the model. Instead, we propose a sampling procedure that initializes a sequence with white noise and follows a Markov chain defined by Langevin dynamics on the global log-likelihood of the sequence. This approach parallelizes the sampling process and generalizes to conditional sampling. Using an autoregressive model as a Bayesian prior, we can steer the output of a generative model using a conditional likelihood or constraints. We apply these techniques to autoregressive models in the visual and audio domains, with competitive results for audio source separation, super-resolution, and inpainting.