From discrete-time policies to continuous-time diffusion samplers: Asymptotic equivalences and faster training
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866908812339838976 |
|---|---|
| author | Berner, Julius Richter, Lorenz Sendera, Marcin Rector-Brooks, Jarrid Malkin, Nikolay |
| author_facet | Berner, Julius Richter, Lorenz Sendera, Marcin Rector-Brooks, Jarrid Malkin, Nikolay |
| contents | We study the problem of training neural stochastic differential equations, or diffusion models, to sample from a Boltzmann distribution without access to target samples. Existing methods for training such models enforce time-reversal of the generative and noising processes, using either differentiable simulation or off-policy reinforcement learning (RL). We prove equivalences between families of objectives in the limit of infinitesimal discretization steps, linking entropic RL methods (GFlowNets) with continuous-time objects (partial differential equations and path space measures). We further show that an appropriate choice of coarse time discretization during training allows greatly improved sample efficiency and the use of time-local objectives, achieving competitive performance on standard sampling benchmarks with reduced computational cost. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_06148 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | From discrete-time policies to continuous-time diffusion samplers: Asymptotic equivalences and faster training Berner, Julius Richter, Lorenz Sendera, Marcin Rector-Brooks, Jarrid Malkin, Nikolay Machine Learning We study the problem of training neural stochastic differential equations, or diffusion models, to sample from a Boltzmann distribution without access to target samples. Existing methods for training such models enforce time-reversal of the generative and noising processes, using either differentiable simulation or off-policy reinforcement learning (RL). We prove equivalences between families of objectives in the limit of infinitesimal discretization steps, linking entropic RL methods (GFlowNets) with continuous-time objects (partial differential equations and path space measures). We further show that an appropriate choice of coarse time discretization during training allows greatly improved sample efficiency and the use of time-local objectives, achieving competitive performance on standard sampling benchmarks with reduced computational cost. |
| title | From discrete-time policies to continuous-time diffusion samplers: Asymptotic equivalences and faster training |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2501.06148 |