Benchmark
APOSTOLUS position
APOSTOLUS has no clinical trial, no external-cohort result and no auditor signature. Its position is defined by what can be replayed today, and this screen lists that beside what does not exist yet.
What exists
- 12 project records, each with a question, an internal target, sources and stated limits.
- 30 evaluation lanes and 15 manifests locked by SHA-256.
- One result on published cases: 172 of 1,874 for project 19-1, below its target.
- 6 harnesses whose numbers a separate audit program recomputed.
- 46 rows of review by models from another family, including the reviews that did not pass at first.
- Replay commands and file hashes for every published result.
What does not exist yet
- No clinical trial, no new participants and no hospital or laboratory work.
- No result on an independent external cohort.
- No auditor signature on any result.
- No peer-reviewed publication and no released dataset.
- No support, selection or recognised outcome under the public call.
What would change the position
A public benchmark result
A project evaluated on its locked public split, reported with its full denominator.
An external cohort
Approved access to row-level data from another institution, with independently adjudicated labels.
The auditor’s signature
Isolated replay of the frozen inputs, signed separately from the automated verdict.