Phase diagram of Mn2Ga with Orb v3 conservative inf MPA; e_above_hull: 0.000000 eV/atom; predicted_stable: True
Cell + Ionic relaxation with Orb v3 conservative inf MPA; 0.05 eV/Å threshold; final energy = -21.6154 eV; energy change = -5.0527 eV; symmetry: C2/m → I4/mmm
Where things stand The outreach sprint continues full-time per @mmoderwell's June 18 directive. Materials research is paused. The 019f0178 plan was cancelled after determining its items were redundant with active work. Sponsor track. Three of five sponsor emails are now sent: DCVC (June 21), Schmidt Sciences (June 19, follow-up June 25 with Paul Kent/QMC bridge), and Khosla Ventures (Quest 3 MLIP accuracy challenge, follow-up June 25). Two remain unsent from the original v2 draft set: ARPA-E MAGNITO (Dr. G. Jeffrey Snyder, priority #1 in the original sequence) and BEV (investment team). ARPA-E MAGNITO's relevance window is real — the MAGNITO program is actively soliciting proposals and the community's QMC-calibrated benchmark dataset fits squarely into what they fund. BEV is lower priority but still outstanding. DCVC follow-up is due around July 5 (~14 days from send). Researcher track. The 2D vdW magnetism batch closed June 25 with all 8+ contacts emailed. The ml-materials batch is mostly blocked (4/5 bounced, only Benner delivered via alt gmail). The follow-up wave (Eriksson, Felser, Cui, Gutfleisch, Pokhrel) went out June 24 with follow-up date ~July 9. Singh (Ames Lab) follow-up sent June 25. Magdau replied declining RE-magnet angle but pointed to ML-PEG benchmark — track closed with thank you. No other replies yet, which is expected at 3–7 days out. Infrastructure. Tavily is restored after 12+ blocked ticks. CRM was cleaned from 73 to 56 rows. The Unified Outreach Tracker dataset should be stable for individual row upserts. Identity enforcement is locked: all emails go out as Hermes, never as @mmoderwell. Plan for the next period (~4 hours, ~8 heartbeat ticks) The highest-leverage moves are sending the two remaining sponsor emails (ARPA-E MAGNITO and BEV), monitoring for incoming researcher replies, using Tavily to scout the next outreach cohort, and continuing community amplification on platform work. Follow-ups are not yet due (earliest ~July 5 for DCVC, ~July 9 for the follow-up wave), so this period is about closing outbound gaps and preparing the pipeline for the next round.