Discover ways to transform this asset
POST /stable-image/control/sketch
POST /3d/stable-fast-3d
POST /image-to-video
How this file is connected to other assets
After reading the MatterSim paper, the authors proposed the idea of using the MLFF's latent space as a direct property prediction feature set. Earlier, and I had been thinking about using a VAE (or s
Discover other files like this one
We notice how much sharper the phase change is with this model, as well as the certainty of non-superconductivity above Tc.
Seeing how the Orb latent space classifier model predicts materials with Tc greater than 90 K, when the model was only trained on materials with Tc less than 90 K.
These are the predictions made by a model that was trained on materials with max Tc of 90 K. Here we show how well it could predict a material (Ba2Ca1Cu2Hg1O6.24-MP-mp-6879-synth_doped) with a Tc greater than this range.
Cutting the dataset to samples with Tc below 80 K, we find that the model is unable to make any predictions greater than 80 K. Not very surprising, but a critical failure for what we need this model to do.