Two-minute read; links to the underlying papers, model cards and case studies inline.
End-to-end materials discovery loop in production. Antibody developability scoring at 0.83 ROC-AUC on external panels. Cross-modal transfer experiments between MD and DFT now reproducible.
Solid-state battery customer screening dopants for high-conductivity electrolyte compositions. Aether surfaced six novel candidates; three made it to wet-lab confirmation.
No refusal-corpus changes this month.
Where does your active-learning loop bottleneck? Reagent cost, plate throughput, hypothesis generation? Reply.
- (1) Materials active learning under reagent constraints — Holst.
- (2) Antibody developability survey paper — external.
- (3) On the brand refresh — design team.