fig10

Figure 10. Representative physical and computational models for solid-state battery design. (A) Transformer-based CGformer architecture for screening high-entropy SEs. Reproduced with permission from ref.^[181]. Copyright 2025, Cell Press; (B) Machine-learning-assisted phase-field modeling of void-evolution modes in LLZO systems. Reproduced with permission from ref.^[182], Copyright 2025, Wiley-VCH; (C) Screening of cation-doped LLZO compositions thermodynamically stable against Li metal using ML models. Reproduced with permission from ref.^[183]. Copyright 2019, Royal Society of Chemistry; (D) Image-based workflow for cathode microstructure quantification, including phase segmentation, feature extraction, and line-intercept analysis. Reproduced with permission from ref.^[184]. Copyright 2025, Wiley-VCH; (E) Evolution of research paradigms from knowledge-augmented human investigation to autonomous AI-agent-driven discovery. DFT: density functional theory; SE: solid electrolyte; LLZO: Li₇La₃Zr₂O₁₂; ML: machine learning; AI: artificial intelligence.