Docs
Blog

in•1mo

Overtraining a neural network for Tc classification

We notice how much sharper the phase change is with this model, as well as the certainty of non-superconductivity above Tc.

Actions

Discover ways to transform this asset

Image-to-image generation for controlled variations of existing images or sketches

POST /stable-image/control/sketch

Generate 3D assets from a single 2D input image

POST /3d/stable-fast-3d

Generate a short video from an image

POST /image-to-video

Dimensions1644x1032

Size365KB

Extensionpng

Connections

How this file is connected to other assets

1 reference

Material embedding Tc classifier experiments
Doesn't seem to be much of an effect. Though there are more predictions made it isn't helping us find a more concrete Tc for these higher Tc materials because of all the uncertainty around the phase t
in•1mo•updated 25d ago
post

Similar files

Discover other files like this one

Uncertain superconducting state classification

Ag0.1In0.9Te1-MP-mp-2597-synth_doped was tested on our classification model, which failed to find Tc and drop probability after Tc.

High-Tc superconductor state classification

Looking at the model predictions for the 200 K above Tc to get to room temperature, we see consistent and stable predictions for the non-superconducting state.

High-Tc superconductor uncertain state classification

Simulating the material heating up to 300 K, we watch as the probability of the superconducting state gradually and steadily decreases, which makes it hard to choose an exact Tc.

Raising the classification threshold for Tc estimation

By increasing the level of certainty, requiring a higher probability for superconductivity, it's possible to get better results. However, this kind of tuning usually comes at a cost and other materials will suffer poorer results.