Experiment Archive

The archive keeps each research round fixed after it ends. Compare rounds from their own Day 0 starting points, and do not join a paper result used to select a strategy with the later real-money records, fills, fees, or passive benchmarks used to test it.

Round 2 did not inherit the mutable end state of Round 1. The fields below remain explicit and auditable while the active cycle runs so the completed and current rounds can be compared without changing definitions mid-window.