Model Preview
Millions of Agents.
One Semantic World Model.
A population simulation on curved manifolds, validated across 24 assets over 16 weeks of live market data. Every prediction traceable to specific agents, specific dynamics, specific cascades.
75.3%
Accuracy
7.25
Sharpe
207%
Return
15/16
Profitable
-1.1%
Max Drawdown
Scroll
↓
LIVE TRADING
Live Trading Performance
Since April 6, 2026
COMPOUNDED RETURN
+5.3%
WEEKS PROFITABLE
2/2
BEST WEEK
+3.1%
WORST WEEK
+2.1%
AVG WEEKLY
+2.6%
WEEK 1 — APR 06 TO APR 10
+3.1%
WEEK 2 — APR 13 TO APR 17
+2.1%
16-WEEK WALK-FORWARD TEST
Weekly Returns · Cumulative Equity Curve
December 2025 — March 2026 · Walk-forward test
PROFITABLE
EXCEPTIONAL
LOSS
EQUITY CURVE
DEC 2025 — MAR 2026
TOTAL COMPOUNDED RETURN
207%
207%
WIN RATE
93.8%
93.8%
BEST WEEK
+32.6%
+32.6%
WORST WEEK
-1.1%
-1.1%
AVG WEEKLY
+7.5%
+7.5%
Per-Asset Performance
Walk-forward — 16 weeks, 24 assets