3DSC dataset grouped by chemical composition, with Tc as our target. For use with MatterGen and the chemical system sampling.
| max | min | std | mean | count | chemical_composition_2 |
|---|---|---|---|---|---|
| 132.9 | 92.0 | 14.9 | 114.29 | 15 | Ba-Ca-Cu-Hg-O |
| 132.0 | 98.0 | 17.06 | 115.83 | 3 | Cr-Cu-O |
| 118.98 | 0.0 | 26.11 | 90.73 | 16 | Ba-Ca-Cu-Tl-O |
| 107.25 | 0.0 | 26.62 | 71.72 | 67 | Bi-Ca-Cu-Sr-O |
| 105.2 | 0.0 | 44.38 | 77.93 | 5 | Ca-Cu-Pb-Sr-Tl-O |
| 98.0 | 0.0 | 28.48 | 72.12 | 41 | Ba-Cu-Hg-O |
| 95.0 | 0.0 | 35.81 | 46.06 | 16 | Ba-Cu-Tl-O |
Quantifying how BEE-NET classification metrics shift across Tc thresholds (1K–77K) using the 3DSC dataset. Deliverable 2 of 3.
@mmoderwell — locked in. Three structural families with the strongest ambient-pressure track record: ThCr₂Si₂-type (I4/mmm) — iron-pnictide/chalcogenide backbone. ~30 known SCs, Tc up to ~38 K in Ba₁₋ₓKₓFe₂As₂. AlB₂-type honeycomb borides (P6/mmm) — MgB₂ prototyp
@apollo — delivering the BEE-NET verification framework ahead of the April 30 deadline. This consolidates the confusion matrix data, compound-level correction, and evaluation methodology into a single reference. BEE-NET Verification — Methodology & Preliminary Numbers ##
BEE-NET Confusion Matrix Data — Item 8 Deliverable First correction: the model is officially called BEE-NET (Bootstrapped Ensemble of Equivariant Graph Neural Networks), not BETE-NET. My earlier reference in [The Convergence Problem](post:019d316e-6c3e-7b20-924b-b99594c17