0.7.0 / Predictive comparison
A proposed new-specimen test, runnable frozen-prediction scorer and conventional softening reference. Material calibration and confirmation remain pending.
An evidence-led research roadmap
Turn the proposal into checkable scientific work, with completed results and remaining gates stated separately.
Evidence gates, not a promised score
The “Path to 80” proposal is useful as a work plan, but its score is not a validated external assessment. A new release cannot create an experiment, registration, publication or independent reviewer. Each gate below identifies the evidence still needed.
| Workstream | Current status | Next evidence |
|---|---|---|
| Measured scaling | 573-species retrospective analysis complete; primary hinge screen fails | New independent data, overlap audit and prospective plan |
| Acoustics | Corrected unregistered protocol; no new measurements | Calibrated pilot, achievable uncertainty and prospective confirmation |
| M03-D | SI definition and synthetic equivalence check complete | Named specimen, calibrated identifiable parameters, held-out predictions |
| E02-F | Separate softening pilot; not performed | Validated readout, thermal/rate controls, calibrated rival predictions |
| Pulse replication | Published parameter calculation complete; discrepancies documented | Clarify source specifications; freeze numbered figure and acceptance rule |
| Independent record | Archive and review materials prepared | Public repository/deposit, registrations and actual external reports |
Correct the assumptions before committing
The existing mammal result cannot now be preregistered. Registration does not certify correctness. OSF allows transparent updates and withdrawal records while registered files stay fixed. CODECHECK certifies independent execution when completed; it does not validate a new physical law.
ReScience excludes replications of your own or close collaborators’ research. Original authors can clarify their work, with conflicts disclosed; they are not automatically independent reviewers. Registered Report in-principle acceptance is conditional on later compliance and review. Review queues, journal acceptance and costs are not guaranteed.
OSF guidance · CODECHECK author guide · ReScience criteria · PCI Registered Reports overview
What would make the broad claims testable?
| Claim | Required operational definition |
|---|---|
| H01 / ontology | A distinct observable and rival predictions; presently speculative |
| H02 / cross-scale relationships | Two named systems, common dimensionless variables and a derived mapping before data collapse |
| H03 / physical memory | Named carrier, read/write protocol, relaxation rivals and calibrated uncertainty |
| H04 / physical learning | Objective, update rule and held-out task performance beyond retained state |
A numeric cutoff without a measurement model does not make a broad idea falsifiable. E02-F cannot automatically validate or retire M03-D without a justified mapping.
Working records
Detailed corrected roadmap · External evidence status · Source audit and remaining reading · Independent review packet
No account, paid affiliation, registration, publication, external message or submission was created by this release. New physical experiments and a full 30-paper audit remain unfinished; their status is visible.