Choose one problem to review.
Specific requests for replication, criticism and correction. Reviewing a model does not require accepting the wider framework.
R01 / Check the energy derivation
Question: Does the model include work from changing stiffness? Derive dE/dt for E = (p² + kq²)/2 and compare the numerical ledger, including signed parametric work.
Useful output: a derivation, a minimal failing parameter set or convergence comparison. Include version and time step. This subsystem ledger cannot establish total device efficiency.
Read the derivation →R02 / Challenge the memory baseline
Question: Which conventional model should the proposed physical model have to beat? Identify the observable, confounds, reset procedure and an independent test split.
Useful output: a lawful dataset and methods record, a baseline specification, or an identifiability objection. A fit alone is insufficient.
Read the flagship protocol →R03 / Reproduce a result
Question: Can another person reproduce a packaged synthetic result without changing inputs?
Useful output: exact command, runtime, dataset hash, output and discrepancy. Synthetic reproduction checks computation, not physical truth.
Get the versioned package →How review becomes public
Send the request ID with evidence and alternative explanations. The form grants private review permission only. Public quotation, attribution and attachments require separate specific permission. Critical and supportive reviews face the same standards; response and acceptance are not guaranteed.
Submit a review →