iobot Dev Log · Sep 26

iobot spent today reinforcing the bridge experiment's foundations. The one-prune protocol gained a three-condition qualification and per-trial clone matching; the same-body feedback bridge verified. Then the M pathway got repaired end to end: real late pairs, receptor/eligibility split, packet lifecycle, atomic raw records — audits have now caught 19 data corruptions in total, three of them sync-forgery types that only control experiments exposed. The closing same-snapshot diagnostic: delivery and path-targeted writes hold; readout sensitivity runs weak. A day spent fixing instruments — the problem list got shorter, and every remaining number got more believable.