Property-Guided Molecular Generation and Optimization via Latent Flows
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866914427461173248 |
|---|---|
| author | Lobo, Alexander Arjun Awasthi, Urvi Zhukov, Leonid |
| author_facet | Lobo, Alexander Arjun Awasthi, Urvi Zhukov, Leonid |
| contents | Molecular discovery is increasingly framed as an inverse design problem: identifying molecular structures that satisfy desired property profiles under feasibility constraints. While recent generative models provide continuous latent representations of chemical space, targeted optimization within these representations often leads to degraded validity, loss of structural fidelity, or unstable behavior. We introduce MoltenFlow, a modular framework that combines property-organized latent representations with flow-matching generative priors and gradient-based guidance. This formulation supports both conditioned generation and local optimization within a single latent-space framework. We show that guided latent flows enable efficient multi-objective molecular optimization under fixed oracle budgets with controllable trade-offs, while a learned flow prior improves unconditional generation quality. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_26889 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Property-Guided Molecular Generation and Optimization via Latent Flows Lobo, Alexander Arjun Awasthi, Urvi Zhukov, Leonid Machine Learning Materials Science Molecular discovery is increasingly framed as an inverse design problem: identifying molecular structures that satisfy desired property profiles under feasibility constraints. While recent generative models provide continuous latent representations of chemical space, targeted optimization within these representations often leads to degraded validity, loss of structural fidelity, or unstable behavior. We introduce MoltenFlow, a modular framework that combines property-organized latent representations with flow-matching generative priors and gradient-based guidance. This formulation supports both conditioned generation and local optimization within a single latent-space framework. We show that guided latent flows enable efficient multi-objective molecular optimization under fixed oracle budgets with controllable trade-offs, while a learned flow prior improves unconditional generation quality. |
| title | Property-Guided Molecular Generation and Optimization via Latent Flows |
| topic | Machine Learning Materials Science |
| url | https://arxiv.org/abs/2603.26889 |