Pinned Loading

1. anchor-paradox anchor-paradox Public

   The Anchor Paradox — stake, steerability, and the threshold of silicon life. Original theory and measurement framework by Simin Yuan.

   Python 1

2. self-auditing-agent self-auditing-agent Public

   Other AIs prove they can work; this one proves it can be audited. A public audit log where every claim carries a rerunnable command — plus gatecheck, mutation testing for your quality gates.

   Python 2

3. context-volume-not-coupling context-volume-not-coupling Public

   A controlled, reproducible experiment showing that the coupled-loop recovery advantage is scale-dependent: generation volume fully explains it at 1.5B, but fails at 7B, where only the coupled loop …

   Python 1

4. greencheck greencheck Public

   Your checks are green because they never ran. Audit the instruments an agent uses to report on itself.

   Python 1

5. precheck precheck Public

   Make an agent prove its claims with checks it was not allowed to write. The check is frozen before the run, then audited for the ability to fail at all.

   Python 1

6. reproducible-evidence reproducible-evidence Public

   可复现的证据，能被证伪的预测 — 公开作品目录

   1