Skip to content

Paper: the transport ladder, the disability gate, and a gate amended by its own evidence#166

Merged
MaxGhenis merged 1 commit into
masterfrom
paper-mega
Jul 12, 2026
Merged

Paper: the transport ladder, the disability gate, and a gate amended by its own evidence#166
MaxGhenis merged 1 commit into
masterfrom
paper-mega

Conversation

@MaxGhenis

Copy link
Copy Markdown
Contributor

Carries paper/paper.qmd through every merged wave since the gate-2b pass (#149) — the gate-2c ladder and verified pass, the disability gate, the transport gate and its self-amendment, the interim benefit seam, the deployment infrastructure, and a full tally recomputation. Prose drafted in the paper's voice; every number verified from merged artifacts, PR bodies, issue-#42 registrations/gradings, gates.yaml, and runs/*.json.

Sections added / changed

Verification

  • Frozen sections byte-identical — all 54 deleted lines are inside the six intended edit regions; the YAML front matter, Introduction–Scoring, gate-1/2a/2b mechanics, @sec-surface, Related work, Status, and References are untouched.
  • Always-runnable suite green inline (worktree, CI-equivalent): 1029 passed, 41 skipped, 1466 deselected — including test_paper_figures (the three committed SVGs still rebuild byte-for-byte).
  • uvx black --check -l 79 . and uvx ruff check . clean; no .py, no committed artifact, and no gates.yaml block changed, so no tier refresh.

🤖 Generated with Claude Code

…by its own evidence

Carry the paper through every merged wave since the gate-2b pass (#149).

Add the gate-2c candidate ladder and its independently verified pass
(#155/#157, verification 4950580430): two candidates — the shortest of the
family ladders — candidate 1 failing 1/5 with the assortative diagonal held
35/35 on the first attempt and the gate bound entirely on the earnings-blind
first-marriage cells the registration had named in advance, candidate 2's
one marginal-preserving modifier (byte-carry 2,500 cells at 0.0) passing 4/5.
With it, every gate-2 tranche is passed: all four demographic gates locked
AND passed (memo 4950584695).

Add the disability gate (M4): floors #153 (round-1 restoring the occupancy-
stock teeth via mixed-k), lock #159 (the start-wave weight-definition fix),
candidate 1 passing 5/5 (#161) — the program's first five-of-five — verified
with a both-constructions gap adjudication (4951142927), the estimand always
the PSID self-report, never an SSA DI level (seven concept deltas).

Add the transport gate (W1): design #151, floors #154 with the owned #158
tier-manifest process violation, lock #160, candidate 1 failing 0/5 (#162,
C2 swap realised / C1 non-reversal robust / earnings not degenerate),
forensics 1 (#163, the entry-state lever and the Q4 gate-design finding),
and amendment 1 in full ceremony (#164/#165) — the program amending its own
structurally-unpassable gate, all ten family-B cells demoted on committed
evidence with anchors retained verbatim and the failed candidate left failed.
This is the fourth recurring referee class: a gate can itself be wrong, and
the same ceremony catches that too, after the lock.

Add the interim benefit seam (#156: gross $41.35B -> net $36.02B, offsets
enumerated, 4/4 expectations, engine-vs-transport gap 6e-8), stating the
PolicyEngine-executes / Axiom-cross-validates division; and the deployment
infrastructure (contract identity, env sidecars, a generic K=20 evaluator
reproducing the three committed gate verdicts; the household candidate-9
registry port).

Recompute the tally per the paper's conventions: forty-two registered runs
across six gates (13/16/9/2/1/1), thirty-seven failures against five passes,
six gates locked with FIVE ratified amendments (two gate 1, two gate 2, one
transport), ten forensics rounds, six external-anchor replications, and fifty
forecasts registered with forty-nine graded — the one open registration the
transport gate's second candidate. Frozen sections byte-identical.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
@vercel

vercel Bot commented Jul 12, 2026

Copy link
Copy Markdown

The latest updates on your projects. Learn more about Vercel for GitHub.

Project Deployment Actions Updated (UTC)
social-security-model Ready Ready Preview, Comment Jul 12, 2026 8:04pm

Request Review

@MaxGhenis MaxGhenis merged commit ab1e4dc into master Jul 12, 2026
7 checks passed
MaxGhenis added a commit to PolicyEngine/populace.dev that referenced this pull request Jul 12, 2026
… gate amended by its own evidence (#42)

Deploy the rendered paper from PolicyEngine/populace-dynamics#166 (merged),
in the pattern of #40 / #34.

- dynamics/web/index.html <- paper/paper.html
  (1,391,551 -> 1,422,260 bytes; SHA-256
  366b86fc433a9d7bafb819a7cea142b00d3d7193bd2a26f02b84647d66888e7e)
- dynamics/populace-dynamics.pdf <- paper/paper.pdf
  (277,783 -> 307,023 bytes; SHA-256
  30db19869f7a1327772f26006cfb515648b5d2febe174ae7664983738860de20)

Rendered with Quarto 1.9.36 (the same toolchain as the prior sync). The
wrapper dynamics/paper.html is unchanged, exactly as #40.

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant