Verification/Evidence review
Free Lab

Contributor terms

Your choices

Scoring alone does not retain your raw trace for a benchmark. Publishing a result stores a scorecard, declared scope, input fingerprint and consent receipt. It does not publish raw trajectories. Joining the public board is a separate choice.

Optional benchmark contribution

By explicitly selecting the benchmark option, you confirm that you have the right to contribute the trace and grant the deployment operator a non-exclusive, royalty-free permission to retain, inspect, reproduce and transform it for verification research, calibration evaluation and reproducible benchmark artifacts. You retain ownership. This is not permission to train or tune your policy.

Do not submit confidential data, personal information, export-restricted material or data whose source terms prohibit this use. You are responsible for source permissions. This draft does not claim that any upstream dataset license permits redistribution.

Visibility and retention

Demo public links expire after seven days. Expiry removes public access; it is not proof of physical deletion. Consented benchmark contributions remain until the operator removes them under its retention process. Free hosting may reset all ephemeral state. Export evidence you need to keep. No production retention assurance is offered.

Attribution and contact

The current beta uses anonymous demo keys, not verified email accounts. It does not collect contributor biographies or invent community attribution. The deploying organization must publish its identity, contact and withdrawal process before inviting private or sensitive contributions.

Limits

Results use DRAFT gates and scoped measurements. They are not legal advice, whole-system safety findings, or evidence of external acceptance. Public submissions may be excluded for duplication or incompatible scope. No model is automatically repaired.

Back to Free Lab