RISCONAPOSTOLUS

Methods & About

Research Ledger

What was registered, checked, run, locked or corrected, and by which script. Corrections stay in the ledger beside the entries they correct.

Entries for a harness are written by the session that ran it. The site assigns the entry numbers in date order.

25 entries5 corrections
EntryDateKindWhat was doneScript or record
L-252026-11-14CorrectedCorrection: cross-model review found that the DP-SGD aggregate added no noise for an empty batch, that the baseline generator failed on an empty data set, and that the specification tied the sampling rate to the private row count. All were fixed with tests; verdicts did not change.docs/review/SYNTH8-DISPENSE18-AJA.md
L-242026-11-14RunWrote the 8-1 specification and ran its harness on a synthetic multi-omics world. A separate audit recomputed 860 checks and passed four gates. The private baseline generator passed the privacy gate but kept 91.4% of rank agreement, so its release was blocked. No real cohort was run.synth8/orchestrate.py · synth8/audit.py
L-232026-11-14CorrectedCorrection: review found that the draft kernel treated a missing audit ledger as a successful commit, that the drift monitor reused a fixed calibration table, and that out-of-range vision values could pass. The kernel now fails closed, the monitor uses online p-values and inputs are range-checked; verdicts did not change.docs/review/SYNTH8-DISPENSE18-AJA.md
L-222026-11-14RunWrote the 18-2 specification and ran its audit kernel on simulated vision outputs and abstract orders. Of 4,500 injected errors 0 passed automatically; 1,859 of 3,000 correct fills did. A separate audit replayed every decision. No image inference was run.dispense18/orchestrate.py · dispense18/audit.py
L-212026-11-14CheckedAudited the 16-2 and 16-3 run with a separate program that does not import the solvers: 254 exact binomial bounds, the primary run recounted from per-patient rows, five verdicts re-derived from trajectories and 21 AUROC figures recomputed from per-case scores. Four gates passed.cog16/audit.py
L-202026-11-14RunRan the 16-3 cross-cohort evaluation on synthetic multi-site cohorts: both train/test directions, three harmonisation modes, label-permutation and leakage controls, and 10 seed replications.cog16/orchestrate.py
L-192026-11-14RunRan the 16-2 virtual-cohort Monte Carlo: 1,000 synthetic patients per run, three seeds, 14 distribution shifts, baselines, fault injection and five protocol cases. 962 met the convergence and consecutive-error criteria in the primary run.cog16/orchestrate.py
L-182026-11-14CheckedRecomputed the audit record from the stored result and the frozen input hashes and checked four gates. The check command reproduces the audit file byte for byte.acmg19/summarize.py
L-172026-11-14CorrectedAfter cross-model review, removed the option to admit literature findings without the second reader and split checking from admission. Findings enter the ledger through one function that requires the second reader’s verdicts.acmg19/literature.py
L-162026-11-14CorrectedAfter cross-model review, removed the argument that let a caller hand the audit its own ledger check. The gate now verifies every ledger itself and works only on the verified copy.acmg19/acmg_audit.py
L-152026-11-14RunRan a language-model curator on 24 held-out variants under two instructions, and a literature reader on 96 sampled variants with a second reader from another model family. The stored answers are audited again on every evaluation.acmg19/model_runs.py
L-142026-11-14RunClassified 8,667 expert-classified variants and ran 480 simulated patients through the agents and the audit gate, then injected 9,493 unsupported claims and 360 ledger edits. Repeated runs give a byte-identical result file.acmg19/evaluate.py
L-132026-11-14LockedFroze six input files and recorded their SHA-256: the ClinGen expert classifications and panel VCF, the annotation store, the simulated cases and their VCF archive, and the literature text windows.acmg19/build_inputs.py
L-122026-11-14RegisteredFixed the 40-gene panel and split it into 20 development and 20 held-out genes by the SHA-256 order of the gene symbols. Rules are adjusted on the development genes only.acmg19/fetch.py
L-112026-11-14BuiltRebuilt the site as IATHEON screens: one screen for each menu item, in the same nine sections, and linked the 12-1 to 12-4 records to their specifications and evidence.scripts/build_site.py
L-102026-11-14CorrectedCorrection: a state-machine property test found that a session already in a terminal state could repeat its notification after a policy fault. Effects now fire only on entering a state, with a regression test and an audit rule.prevent12/hypothesis_spec
L-092026-11-14RunWrote design specifications for 12-1 to 12-4 and ran their reference implementation on fixed synthetic inputs. A separate audit program recomputed the numbers by other methods and passed four gates.prevent12/orchestrate.py · prevent12/audit.py
L-082026-11-14BuiltPublished the bilingual site with English at the root and Korean under /kr/, using the IATHEON header, full menu and RISCON Labs family line.scripts/build_site.py
L-072026-11-14RunEvaluated 19-1 on 1,874 published cases. 172 had the reported gene in the top five (9.18%), below the 90% target. The replay matched byte for byte, including a run with the network blocked.gate1/rare19/replay.sh
L-062026-11-14LockedFixed the evaluation data, reference labels and splits for all twelve projects. Fifteen case or group manifests were locked by SHA-256.gate1/assemble.py
L-052026-11-14CheckedReviewed the code attached to 5-1 and 18-2 and recorded which claims it does not yet support.docs/ATTACHMENT-REVIEW.md
L-042026-11-14RunRan the 19-1 synthetic runner and its separate audit program on four synthetic VCF cases. The audit passed four gates; two of four cases reached the top five.rare19/orchestrate.py · rare19/audit.py
L-032026-11-14RunRan the common synthetic suite: 12 projects, 14 checks and 78 fixed cases, each compared with its expected value.engine/run_audit.py
L-022026-11-14CheckedChecked the twelve project names and requirements against Ministry of Health and Welfare Notice 2025-961 and its proposal requirements, page by page.docs/SOURCE-REGISTER.md
L-012026-11-14RegisteredRegistered twelve projects with their internal targets, 28 public sources and access terms. APOSTOLUS keeps its work and records apart from other RISCON Labs research sites.site-data/tracks.json

© 2026 RISCON Co., Ltd.Privacy PolicyTerms of UseCookie SettingsContact

RISCON Labsrisconlabs.com
RISCONAPOSTOLUSIndependent Health and Medical Research