How far should an embodied system change after unexpected experience? Research OS, Mismatch Lab, and Builder Lab by Seonhye Gu.
Updated: 2026-07-30
Scientific phase: confirmatory complete
Writing phase: venue-neutral manuscript and supplement v1.1 built; author, affiliation, and independent-research declarations added; venue conversion remains
The frozen model-order confirmatory passed every conjunctive criterion. The core evidence package is complete enough for a full paper draft.
| Milestone | Status |
|---|---|
| Isaac persistent-drift anchor | Complete and passed |
| Shared-process pilot v0.2 | Invalidated and excluded |
| Process-isolated pilot v0.3 | Complete; calibration only |
| Confirmatory preregistration v1.0 | Frozen before fresh data |
| Confirmatory VESSL execution | Complete: 400/400 Ep2 and 20/20 retention cells |
| Confirmatory validity | Passed |
| H1 prediction | Passed |
| H2 fixed-horizon behavior | Passed |
| H3 static retention | Passed |
| H4 drift/control gate validity | Passed |
| Manuscript v1.1 | Submission-oriented review draft and PDF built |
| Supplement v1.1 | Complete review draft and PDF built |
| Bibliography | Citation audit complete; BibTeX added |
| Figures and CSV tables | Generated and hash-manifested |
| Author and disclosure pass | Seonhye Gu; independent personal research; affiliation limitation stated |
Exact report:
RESULTS.md.
| Item | Identity |
|---|---|
| Preregistration tag | paper002-model-order-confirmatory-v1.0 |
| Frozen run source | bea5a6ce78755db6f94b9cb74f6e841e3ebf6f01 |
| Raw artifact commit | 73a7e16 |
| Config SHA-256 | 9cbca0202f784a3a60569dbcf6f36e31a484c95393cb72ef5a030909e9e19635 |
| Execution isolation | Fresh Isaac process per seed-arm-condition |
The positive result is scoped to a prepared zero-order-to-constant-velocity expansion in one Isaac surgical-reach target-dynamics family. Hardware, tissue-contact, autonomous representation invention, and clinical claims remain open.