Salvato in:
| Autori principali: | , , , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | https://arxiv.org/abs/2605.14759 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866911715157868544 |
|---|---|
| author | Liu, Nian Kazeev, Nikita Dale, Stephen Gregory Maevskiy, Artem Zeng, Yuwei Kubo, Ryoji Huang, Pengru Laurent, Thomas LeCun, Yann Novoselov, Kostya S. Bresson, Xavier |
| author_facet | Liu, Nian Kazeev, Nikita Dale, Stephen Gregory Maevskiy, Artem Zeng, Yuwei Kubo, Ryoji Huang, Pengru Laurent, Thomas LeCun, Yann Novoselov, Kostya S. Bresson, Xavier |
| contents | De novo crystal generation seeks to discover materials that are not merely realistic, but also stable and novel. However, most existing generative models are trained to maximize the likelihood of observed crystals, which encourages samples to stay close to known materials yet not necessarily align with the criteria that matter in discovery. Our empirical analysis shows that current crystal generative models exhibit a clear conflict between stability and novelty: samples near the observed distribution tend to retain stability but offer limited novelty, whereas samples farther from it often lose stability rapidly. This suggests that the useful region for discovering crystals that are both stable and novel is extremely narrow. To move beyond this limitation, we introduce Crys-JEPA, a joint embedding predictive architecture for crystals that learns an energy-aware latent space preserving formation-energy differences. In this space, stability assessment can be reformulated as an embedding-based comparison against accessible training crystals, reducing the reliance on expensive energy evaluation and task-specific external references. Building on Crys-JEPA, we further develop a screening-and-refinement pipeline that identifies promising generated crystals and reintroduces them to refine the generative model. On MP-20 and Alex-MP-20 datasets, we achieve improvements over baselines up to 53.8% and 72.7% on V.S.U.N. metric, respectively. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_14759 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Crys-JEPA: Accelerating Crystal Discovery via Embedding Screening and Generative Refinement Liu, Nian Kazeev, Nikita Dale, Stephen Gregory Maevskiy, Artem Zeng, Yuwei Kubo, Ryoji Huang, Pengru Laurent, Thomas LeCun, Yann Novoselov, Kostya S. Bresson, Xavier Machine Learning De novo crystal generation seeks to discover materials that are not merely realistic, but also stable and novel. However, most existing generative models are trained to maximize the likelihood of observed crystals, which encourages samples to stay close to known materials yet not necessarily align with the criteria that matter in discovery. Our empirical analysis shows that current crystal generative models exhibit a clear conflict between stability and novelty: samples near the observed distribution tend to retain stability but offer limited novelty, whereas samples farther from it often lose stability rapidly. This suggests that the useful region for discovering crystals that are both stable and novel is extremely narrow. To move beyond this limitation, we introduce Crys-JEPA, a joint embedding predictive architecture for crystals that learns an energy-aware latent space preserving formation-energy differences. In this space, stability assessment can be reformulated as an embedding-based comparison against accessible training crystals, reducing the reliance on expensive energy evaluation and task-specific external references. Building on Crys-JEPA, we further develop a screening-and-refinement pipeline that identifies promising generated crystals and reintroduces them to refine the generative model. On MP-20 and Alex-MP-20 datasets, we achieve improvements over baselines up to 53.8% and 72.7% on V.S.U.N. metric, respectively. |
| title | Crys-JEPA: Accelerating Crystal Discovery via Embedding Screening and Generative Refinement |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2605.14759 |