Bibliographic audit reports performance metrics across public preprints, highlighting an absence of verified clinical superiority for the evaluated AI scaffold.
A catalogue of twelve distinct public research works (preprints) by R. E. W. Kho, updated on 13 September 2026, version 0.2. The completed radiology version replaces the earlier version within the same work; it is not a thirteenth research work. This index is a finding aid, not an additional research study. Repository publication and a DOI do not establish journal acceptance, peer review, scientific validity, clinical benefit or legal significance. The completed radiology v0.2 reports all 210 stock-control answers: Athena 191/210 (90.95%) and stock Astra 193/210 (91.90%). Stock scored higher descriptively. The non-resource-matched exploratory comparison does not establish an Athena advantage or a scaffold-specific causal effect. Identical underlying checkpoints and full-session compute were not independently verified; the live workflow was only partially operationalised. The public, exposed questions were presented in overlapping image bundles, not an official held-out benchmark protocol. No radiologist-reader experiment or patient outcome was measured. The conditional impact model is hypothetical and includes zero benefit and harm; it is not evidence of clinical safety or lives saved. The psychiatry note corrects unsupported clinical-impact extrapolations in legacy manuscripts; it does not validate every remaining claim in them. The radiology entry links to DOI 10.5281/zenodo.22730550. The preceding radiology version remains at DOI 10.5281/zenodo.22666681. The original catalogue is preserved at Zenodo record 22666823, DOI 10.5281/zenodo.22666823, version 0.1, dated 9 September 2026. This v0.2 updates that dated finding aid and retains the other eleven research entries. It does not replace the historical files or rewrite the earlier radiology snapshot. At this 13 September 2026 checkpoint, the only verified journal submission is "When the rater moves the threshold: a measurement audit of Bodea's synthetic consciousness score", to Cognitive Processing, dated 24 August 2026. The original Springer Nature confirmation verifies submission to the publisher's workflow; the manuscript and submission history identify the journal. The receipt does not establish acceptance, completion of peer review or a current editorial decision. New journal dispatch is ongoing; no new submission receipts have yet been verified. This status note does not add a work to the twelve-entry catalogue. Prepared with AI assistance. The author is developing the Athena system and has an interest in its possible commercialisation. Original catalogue text: CC BY 4.0, R. E. W. Kho, 2026. Linked records retain their own licences. PDF, Markdown and a self-contained HTML view are supplied.
No takes yet. Share an insight, caveat, or question.
R. E. W. Kho (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: