TargetSpace Labs does not own the measurement science it applies. The methodology for evaluating personal and organizational world models — sealed prospective forecasting, explicit baselines, calibration gates, specificity controls — is defined and maintained as an open research standard at targetspace.org. This page explains how the two relate.
The open benchmark defines what counts as evidence of target-specific predictive skill. TargetSpace Labs applies that definition, under the same controls, for teams that need a managed evaluation of their own systems.
The public home of the TargetSpace benchmark: the paper, the evaluation protocol, the open-source harness, task and forecast schemas, and the governance process that decides how the standard evolves. It publishes the benchmark's pilot status openly — the project is pre-pilot, with no human-subject results claimed. It serves researchers, benchmark contributors, and anyone who wants to scrutinize, replicate, or extend the methodology.
The commercial arm of the open TargetSpace benchmark. TargetSpace Labs runs managed prospective evaluations for companies and research labs — design, sealing, baselines, scoring, and a standardized TargetSpace Report — and operates the pilot partner program. It serves teams whose products claim to know, remember, or personalize, and who need that claim tested rather than asserted.
The protocol, the paper, and the harness are public. Nothing in a commercial engagement depends on material you cannot inspect.
Commercial evaluations never weaken the protocol. Where a commercial constraint conflicts with the open protocol — a shortened observation window, a restricted baseline, an outcome that cannot be sealed as specified — the TargetSpace Report discloses the deviation explicitly, so readers can weigh the result against the standard rather than an approximation of it.