Playable brain: deterministic model, real spiking network, per-sentence math, calibration (which caught an inverted trust score) - #92

Merged

slavazeph-coder merged 14 commits into

mainfrom

claude/simulator-playground-strategy-6dl25j

Jul 30, 2026

Commits on Jul 30, 2026

Make Mind-Hack Autopsy a playable arcade lab and add attractor evidence export

slavazeph-coder
and
claude
committed
Address review: clipboard guard, evidence defaults, DOM-attached download link

slavazeph-coder
and
claude
committed
Add shareable challenge links to the Content Reaction Lab

slavazeph-coder
and
claude
committed
Make the brain simulation deterministic, parameterized and measurable

slavazeph-coder
and
claude
committed
Show the per-sentence math behind every content score

slavazeph-coder
and
claude
committed
Add ordinal calibration harness — and record that trust ranks backwards

slavazeph-coder
and
claude
committed
Fix the inverted trust signal that calibration exposed

slavazeph-coder
and
claude
committed
Add a real LIF spiking network validated against Brunel (2000)

slavazeph-coder
and
claude
committed
Add Defend the Brain — the arcade's first machine-checked mission

slavazeph-coder
and
claude
committed
Add inter-annotator agreement and verifiable run proofs

slavazeph-coder
and
claude
committed
Add external-corpus evaluation and the annotation rubric

slavazeph-coder
and
claude
committed
Put the spiking network on screen, pay XP for real wins, close the trust gap

slavazeph-coder
and
claude
committed
Give the scores a meaning that is actually true, and show the calibration

slavazeph-coder
and
claude
committed
Close the verification loop and document what was measured

slavazeph-coder
and
claude
committed