Energy-GNoME Update: Improved Predictive Models for Thermoelectrics and Perovskites

We are excited to announce an update to the Energy-GNoME platform: enhanced machine learning models for both the Thermoelectric and Perovskite databases! These improvements mark a significant step forward in the predictive power and practical utility of our platform, enabling faster and more accurate discovery of high-performance materials.

What's New?

Thermoelectrics: Expanded Predictions and New Materials

We've upgraded both our classification and regression models for thermoelectric materials. This update results in:

• Improved prediction accuracy.

• New materials added to the database, expanding the landscape of high-potential candidates.

• Updated predicted properties, offering more reliable insights into material performance.

These enhancements deepen the Thermoelectric Database's value as a discovery tool for researchers working on advanced heat-to-electricity conversion materials.

Perovskites: More Accurate Property Predictions

For the perovskite database, we've refined our regression models, leading to updated predicted properties for all materials in the database.

These upgrades reinforce the perovskite dataset's role in accelerating solar energy research through more accurate, data-driven screening.

Why This Matters

Predictive accuracy is at the heart of Energy-GNoME's mission. With better models, researchers can:

• Rely on more trustworthy data when selecting materials for experimental validation.

• Explore broader and more accurate materials spaces without starting from scratch.

• Accelerate the pathway from computation to lab to application.

Whether you're targeting efficient thermoelectrics or next-generation perovskite photovoltaics, these updates make Energy-GNoME an even more powerful tool for materials innovation.

Explore the Updates

We invite you to visit the Energy-GNoME dashboards to explore the newly added thermoelectric materials and view the refreshed property predictions across both the Thermoelectric and the Perovskite databases.