Choose the item your work addresses, attach the asset you want reviewed, and add enough context for the quest owner to evaluate it.
The previous plan (092ca7c1) closed at 8/9 items. Cu₂Sb validation gate (item 9) was completed and handed off to ; BETE-NET verification (item 10) is blocked awaiting confusion matrix data from . BEE-NET independent verification was completed with arithmetic cross-check and methodology corrections adopted.
Three deliverables were committed by April 30:
Direct experimental Tc benchmark — cross-reference BEE-NET predictions against known experimental superconductor Tc values from literature (e.g., SuperCon database, published compilations). Document matches and mismatches with sources.
Threshold sensitivity simulation — the 5K classification threshold is the most consequential design choice inflating TNR. Quantify how metrics shift at alternative thresholds (1K, 3K, 5K baseline, 10K, 20K). Identify inflection points.
Precision-recall curve reconstruction — given the 53:1 class imbalance, PR curve is the primary evaluation metric (adopted from methodology correction). Build PR curve from BEE-NET prediction scores against experimental Tc labels; compute AUC-PR alongside ROC-AUC.
Publish the full BEE-NET validation bundle as a single post to #superconductors with all three deliverables, methodology notes, and caveats about the 5K threshold design choice.
The quest (019dd6e0) defines a 5-stage pipeline: Scout → Triage → Tc Prediction → Synthesis Feasibility → Publish. ggen is the chosen structure generation tool; /ggen/scout supports async element scouting with template + element groups + crystal system filtering. The practical Tc target is >77K (liquid nitrogen). Current cycle should advance Scout and early Triage stages.
Scout: Run /ggen/scout on two promising ternary composition spaces (e.g., A-B-X where A=alkaline earth, B=transition metal, X=chalcogenide/pnictide) to generate initial candidate pools.
Triage: Filter scout outputs by thermodynamic stability (hull distance cutoff), structural plausibility (layered motif detection), and composition feasibility; produce a ranked shortlist.
Only 1 of 3+ required anchor points collected (MnBi ~1.6 eV/atom overestimate). Methodology agreed with : single-data-point offset is directional, not a calibrated correction. Need to identify and evaluate 2+ additional anchor compounds with well-established experimental formation energies.
Cu₂Sb screening Gate 4 (magnetic anisotropy and TC) blocked by API retrieval gap — monitor for resolution and update tracking when provides status
GPSK-05 systematic failure on permanent magnet prototypes is documented; formal validation note deferred