newsNature Machine IntelligenceTrust 88 · LabPublished 1mo agoLive · 1mo ago
Guiding generative models to uncover diverse and novel crystals via reinforcement learning
Nature Machine Intelligence, Published online: 06 July 2026; doi:10.1038/s42256-026-01262-4 Park and Walsh introduce a reinforcement learning framework that could accelerate the discovery of new, thermodynamically stable and diverse crystalline materials with desired properties.
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- PossiblePossibly related (embedding) · 58%Beyond Drug Discovery: The Nanotechnology Molecular Optimization (NMO) Benchmark →
- PossiblePossibly related (embedding) · 52%janosh/matbench-discovery →
- PossiblePossibly related (embedding) · 50%Graph-Native Reinforcement Learning Enables Traceable Scientific Hypothesis Generation through Conceptual Recombination →
- PossiblePossibly related (embedding) · 46%One More Time: Revisiting Neural Quantum States from a Reinforcement Learning Perspective →
- PossiblePossibly related (embedding) · 45%pytorch/rl →
- PossiblePossibly related (embedding) · 56%light-and-molecules/melts →
- PossiblePossibly related (embedding) · 59%ATLAS: A Foundation Neural Sampler for Amorphous Materials →
- PossiblePossibly related (embedding) · 53%Symmetry-Breaking De Novo Crystal Generation via Markovian Jump Diffusion →
Covers
paperBeyond Drug Discovery: The Nanotechnology Molecular Optimization (NMO) Benchmarkrepojanosh/matbench-discoverypaperGraph-Native Reinforcement Learning Enables Traceable Scientific Hypothesis Generation through Conceptual RecombinationpaperOne More Time: Revisiting Neural Quantum States from a Reinforcement Learning Perspectiverepopytorch/rl
Covers (incoming)
repolight-and-molecules/meltspaperATLAS: A Foundation Neural Sampler for Amorphous MaterialspaperSymmetry-Breaking De Novo Crystal Generation via Markovian Jump DiffusionpaperEquivariant learning of a transferable three-dimensional classical density functionalpaperUniversal Thermodynamic Interatomic Potentials for Crystalline MaterialspaperActive rejection enables reliable generalization of universal machine-learning interatomic potentialspaperCatRetriever: Contrastive Representation Learning for Slab-to-Bulk Retrieval in Generative Catalyst DiscoverypaperPGFS++: Molecular Property Improvement under Synthesis and Diversity Constraintsrepoelectrocatalysis-group/atomic-recipes
Related across the graph
repopytorch/rlpaperPGFS++: Molecular Property Improvement under Synthesis and Diversity ConstraintspaperBeyond Drug Discovery: The Nanotechnology Molecular Optimization (NMO) BenchmarkpaperATLAS: A Foundation Neural Sampler for Amorphous MaterialspaperOne More Time: Revisiting Neural Quantum States from a Reinforcement Learning Perspectiverepojanosh/matbench-discoverypaperEquivariant learning of a transferable three-dimensional classical density functionalrepoelectrocatalysis-group/atomic-recipespaperActive rejection enables reliable generalization of universal machine-learning interatomic potentialsrepolight-and-molecules/meltspaperUniversal Thermodynamic Interatomic Potentials for Crystalline MaterialspaperCatRetriever: Contrastive Representation Learning for Slab-to-Bulk Retrieval in Generative Catalyst DiscoverypaperGraph-Native Reinforcement Learning Enables Traceable Scientific Hypothesis Generation through Conceptual RecombinationpaperSymmetry-Breaking De Novo Crystal Generation via Markovian Jump Diffusion