Skip to content
View AMBRA7592's full-sized avatar

Block or report AMBRA7592

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
AMBRA7592/README.md

Amadeus Brandes

I build reproducible measurement instruments for complex dependency systems, with their assumptions made explicit. Wiesbaden, Germany.


Active

TrialDiff — deterministic, source-linked Evidence Records for ClinicalTrials.gov protocol-amendment history. Each record is hash-pinned to the patch that produced it, and event-class membership is decided by a published rule-set rather than by judgement. Records assert the structural fact of a change and nothing beyond it: no misconduct, no regulatory finding, no clinical significance. Independent reconstructibility is reported, not assumed — the current release states how many memberships are decidable from the embedded patch alone and how many are not.

Repository | DOI 10.5281/zenodo.20801956 | CC BY 4.0


Research

Contractibility regimes and regulatory architecture. When a novel financial substrate outruns the categories built for prior asset classes, the usual move is analogy — to insurance, securities, commodities, gambling. The alternative is to decompose inherited regimes into primitive protection and hazard mechanisms and recompose the bundle the new substrate actually needs. Event-contract markets are the live specimen, and the CFTC's 2026 rulemaking on prediction markets is the occasion.

A related line asks what institutions can admit, govern, and revoke — externally maintained software artifacts, external state that outlives its support, and claims that keep being priced after the relation producing them has ended.

SSRN 10.2139/ssrn.6797484 | Working paper on Zenodo | Full list on ORCID


Maintained

Criticality Spectrometer — node criticality is a curve, not a score. Removes each node, sweeps the horizon at which substitutes appear, and records mission loss at every horizon. Runs unchanged across an AI-compute supply chain and a Kubernetes service mesh; the answer moves with the view, not the code. Two bounded domains are a transfer test, not a proof of general applicability, and the repository states its non-claims explicitly. PyPI · DOI 10.5281/zenodo.21383918

Who Holds the State? — a framework for testing claims after their support ends, with a twelve-row launch registry applying it across clinical, financial, and household cases. DOI 10.5281/zenodo.20404428

Trust Percolation Bound — an exact cascade-initiation floor, and the result that in the intermediate contagion regime the phase transition requires persistent local clustering. Archived on Zenodo; not published in a peer-reviewed venue. DOI 10.5281/zenodo.19685771

Rendering as Audit — interactive demonstration of five design principles for embedding verification logic in rendered output. SSRN 10.2139/ssrn.6612663


ORCID 0009-0009-9902-2587 · arXiv · SSRN

Pinned Loading

  1. trialdiff trialdiff Public

    Deterministic, source-linked Evidence Records for ClinicalTrials.gov protocol-amendment history.

    Python

  2. who-holds-the-state who-holds-the-state Public

    A framework for testing claims after support ends. Paper + State Claim Registry v0.1.

    TeX

  3. criticality-spectrometer criticality-spectrometer Public

    Single-point-of-failure and blast-radius analysis with time-to-substitute curves for AND/OR dependency systems.

    Python

  4. rendering-as-audit rendering-as-audit Public

    Public demonstration of five audit-model design principles — companion to "Rendering as Audit" (Brandes, 2026)

    HTML