Small Brain. Big Clean.
Fix neural Roomba loops with bounded coverage rewards and measured motor retraining
71db7c7dff335ce1e10fc01b5d6ef4c5ed3f244b7e6789d6bc09fa3ccf01ea1b- Author
- Deata
- Date
- Sep 13, 2026
- Branch
- main
- Parent
- 3700995aa2
14 files changed
- ChangedRoombaExperiment.componentFolderBeforeAfter
- ChangedRoombaExperiment.component/init.luau18.5 KB → 19.4 KBBeforeAfter
- ChangedflyOnlineLearning.moduleFolderBeforeAfter
- ChangedflyOnlineLearning.module/README.md569 B → 1.1 KBBeforeAfter
- ChangedflyOnlineLearning.module/init.luau2.0 KB → 3.3 KBBeforeAfter
- AddedflyOnlineLearning.module/loopFixValidation.json15.1 KBAfter
- AddedflyOnlineLearning.module/regression.luau1.9 KBAfter
- ChangedflySynapticTraining.moduleFolderBeforeAfter
- ChangedflySynapticTraining.module/README.md969 B → 1.5 KBBeforeAfter
- ChangedflySynapticTraining.module/checkpoint.json5.6 KB → 20.4 KBBeforeAfter
- ChangedflySynapticTraining.module/init.luau5.3 KB → 5.3 KBBeforeAfter
- AddedflySynapticTraining.module/navigationRegression.luau1.4 KBAfter
- ChangedroombaHUD.moduleFolderBeforeAfter
- ChangedroombaHUD.module/init.luau6.6 KB → 7.0 KBBeforeAfter