New V-PaSS system improves perovskite solar cell prediction

Researchers have developed the Virtual-Alloy Physics-Aware Semantic System (V-PaSS), a novel architecture designed to improve the predictive accuracy of machine-learning models for perovskite solar cells. Unlike previous models that treat material components as simple labels or rigid structures, V-PaSS incorporates the physical continuity of perovskite solid solutions by interpolating crystal graphs and learning bandgap bowing. It also couples layer sequences, cell polarity, and chemical descriptions using a scientific language model. In evaluations, V-PaSS demonstrated superior performance, achieving the lowest error rate among fourteen tested models and significantly outperforming baselines, particularly on devices with new alloys or unseen transport layers. AI

IMPACT This new system could accelerate the discovery and development of more efficient perovskite solar cells by improving material prediction accuracy.

RANK_REASON The cluster contains a research paper detailing a new scientific model and its performance. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• arXiv
• gradient-boosted baseline
• Hugging Face
• perovskite solar cell
• scientific language model
• Vegard's law
• Virtual-Alloy Physics-Aware Semantic System
• V-PaSS

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →