Forkwise explores how analytical choices affect AI-driven findings. Multiple AI analysts independently choose controls, outlier treatments, and missing-data rules, while Python runs the regressions. A judge compares the results and classifies each finding as robust, fragile, choice-dependent, or showing no effect. See whether your conclusions hold up when reasonable analytical choices change.
