Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This review synthesizes recent advancements in generative models, multimodal learning, and closed-loop workflows for inverse materials design, emphasizing their shift from forward prediction to targeted candidate proposals. The authors evaluate various model classes, including variational autoencoders and diffusion models, and discuss the integration of feasibility constraints and physical priors throughout the design process. Key findings highlight the importance of multimodal learning in creating transferable representations of chemical space and address common failure modes that can hinder effective materials discovery.
Generative models can now propose targeted materials candidates by effectively integrating diverse data modalities and enforcing physical constraints throughout the design process.
Inverse materials design is shifting materials discovery from forward prediction to targeted proposal of candidates that satisfy objectives under physical constraints. Here, we review recent advances in generative crystal structure modeling, multimodal learning, and closed-loop design pipelines for crystalline solids. We survey how modern generators learn chemical-structural priors from large databases to enable controllable sampling of periodic structures, and compare leading model classes including variational autoencoders, normalizing flows, autoregressive formulations, and diffusion models. Particular attention is given to how feasibility constraints and physical priors are enforced across the workflow, through representation choices, training objectives, sampling-time guidance, and post-generation screening and relaxation. We also discuss how multimodal learning fuses diverse materials modalities, including crystal structures, thermodynamic, electronic information, microscopy, spectroscopy, processing context, and scientific text, to construct a more universal, transferable representation of chemical space. In addition, diverse inverse-design strategies are examined, particularly those that integrate conditional generation with latent optimization, Bayesian optimization, reinforcement learning, and active learning. Finally, we highlight recurring failure modes, such as surrogate exploitation, diversity collapse, distribution shift, and the stability-synthesizability gap, and outline discovery-grade evaluation practices based on staged reporting of validity, novelty, uniqueness, stability, and cost.