# Sundog External-Review Tracker

**Purpose.** One surface for every lane waiting on (or cleared by) an external
sanity check. Each lane has a staged reviewer packet; this doc tracks *what's
open, who can answer it, what the owner must do to send, and what clears when it
lands*. The source of truth for the questions is each lane's packet (linked
below); this doc is the dispatch / decision surface, not a re-statement.

**Last updated:** 2026-06-02.

**Status legend:** `DRAFTED` (packet ready, not sent) · `SENT` (awaiting reply) ·
`RETURNED` (verdict in) · `RESOLVED-IN-HOUSE` (no external review needed) ·
`CLOSED`.

> **Synthetic pre-review feedback is logged but does NOT change any lane status.**
> See [Pre-review synthesis feedback](#pre-review-synthesis-feedback-not-external-review).
> The qwen/wolfram synthesis is not a human domain expert; a lane only moves to
> `RETURNED` on a real external reply. Those notes are *noted, not acted on*.

## At a glance

| Lane | Status | Reviewer community | Open Qs | Owner next action |
| --- | --- | --- | --- | --- |
| Navier–Stokes C1 | **SENT** (Titi) + DRAFTED (Olson parallel) | PDE / determining-modes / data-assimilation | 4 | Awaiting Titi reply; send the parallel Olson first-wave note (draft staged) |
| Kakeya | DRAFTED | finite-field incidence / additive combinatorics | 5 | Publish unlinked review URL + permalinks + screenshots, pick reviewer, send |
| Hodge | DRAFTED | algebraic geometer / Hodge theorist | 6 | Fill `[Name]`/`[link]`/`[Your name]`, send |
| Yang–Mills | **SENT** (2026-06-01) | lattice gauge theorist | 7 | Awaiting reply — sent to Dr. Biagio Lucini via LinkedIn DM |
| Riemann | RESOLVED-IN-HOUSE | — | 0 (3 resolved) | None for framing; publish decision is separate |

**The bottleneck is uniform.** The open lanes are *written and waiting*, each
blocked only on "identify a specialist + fill placeholders + send" — to **four
different communities** (no reviewer covers two). The cost is reviewer
identification and dispatch, not drafting. *(Yang-Mills is now out the door — sent
to Dr. Biagio Lucini via LinkedIn DM, 2026-06-01; awaiting reply. The team reports
good peer connections on LinkedIn but no DM replies yet.)*

## The one question underneath all of them

Across reader lanes (Kakeya, Hodge) and empirical lanes (NSE, YM), the
load-bearing call is the same: **is the Sundog framing (regime-2 / body-resistance
/ separation) fenced from overclaim, or is it just ordinary / standard X?** A
"fenced and non-vacuous" in any lane de-risks the portfolio's framing; a "this is
standard / overclaimed" anywhere is the most informative possible negative.

| Review concern | NSE-C1 | Yang-Mills | Hodge | Kakeya |
| --- | --- | --- | --- | --- |
| Overclaim fence (regime-2 / body-resistance bounded?) | Q1–Q2, Q4 | Q2 | Q3 | **Q3 ★** |
| Non-vacuity (more than ordinary / standard / textbook?) | Q4 | Q2 | one-comment (b) | Q1, Q5 |
| Faithful rendering of the established math | Q1–Q2 | Q1, Q3–Q5 | Q1–Q2 | Q1 |
| Methodology soundness (empirical machinery) | Q1–Q3 | Q1, Q3, Q6 | — | Q2, Q4 |
| Pre-reg / anti-p-hunting discipline | implicit | Q7 | — | — |

★ = the lane's own "if you have only one comment" focus.

**Cross-references now live.** The Yang-Mills packet's centerpiece (Q2) rests on
the Faraday Phase 7/8 abelian-bridge note; the Kakeya packet cites
[`CROSS_SUBSTRATE_NOTES.md`](CROSS_SUBSTRATE_NOTES.md) §8. The Faraday / Maxwell
work is load-bearing in two pending reviews.

## Per-lane detail

### Navier–Stokes C1 — DRAFTED (active hold)

- **Under review:** the finite-Galerkin **state-insufficient yet
  control-sufficient** `Phi_K` separation — the strong, provisional regime-2
  witness (currently UNPROMOTED).
- **Load-bearing call (Q4):** a genuine state-reconstruction ↔ control-sufficiency
  separation, or **ordinary data-assimilation / observer behavior** that should
  not be made load-bearing? (Q1: is "state-insufficient on the support" fair for
  the twin-state certificate? Q2: is "control-sufficient on `Phi_K`-fibers up to a
  measure-zero decision surface" fair? Q3: is the held-out look-ahead-max quantile
  a defensible regime-portable objective?)
- **Pre-empted objection:** "isn't this just near-determination?" — the
  objective-overlap discriminator (`proof/PDE_C1_OBJECTIVE_OVERLAP_DISCRIMINATOR.md`)
  shows control-sufficiency does *not* track predictability (Spearman −0.75 / −1.0).
  One honest open puzzle: palinstrophy (predictable but not control-sufficient).
- **Dispatch state:** Titi **sent** (2026-06-02, per the Titi draft Sender Note;
  owner to confirm exact date/channel). **Parallel first-wave Olson note staged**
  (`internal/outreach/NSE_C1_OLSON_EMAIL_DRAFT.md`) — most direct CDA /
  determining-modes fit, recommended next send. Both are individually addressed
  one-to-one sends, not a blast.
- **Attachment (solved):** `internal/outreach/PDE_C1_REVIEW_BUNDLE.pdf` (902 KB,
  the 4 primary docs + 3 run receipts inlined + on-brand closing page; rebuild via
  `node scripts/build-c1-review-bundle.mjs` then `node scripts/render-c1-bundle-pdf.mjs`).
  Verified IP-clean and number-accurate vs the run receipts 2026-06-02. The `.md`
  source bundle remains for anyone wanting plain text. **Templates:**
  [`proof/PDE_C1_EXTERNAL_REVIEW_EMAIL_DRAFT.md`](proof/PDE_C1_EXTERNAL_REVIEW_EMAIL_DRAFT.md)
  (general) + Titi/Olson specifics under `internal/outreach/`. Every headline figure in the packet verified against the run
  receipts 2026-06-02 (942,834 pairs, δ_H 0.0111, ε_K 0.0664, a_mm −0.00079/+0.00058,
  damp 0.3003/0.2688 — all match).
- **Packet:** bundle above, or individually
  `proof/PDE_C1_SEPARATION_STATEMENT.md` (primary),
  `proof/PDE_C1_REGIME_GENERALITY_v1.md`,
  `proof/PDE_C1_KNN_CONVERGENCE_CHECK.md`,
  `proof/PDE_C1_TWIN_STATE_CERTIFICATE.md`,
  [`SUNDOG_V_NAVIERSTOKES.md`](SUNDOG_V_NAVIERSTOKES.md).
- **Gate:** paper-facing use + promotion of C1; public surface.

### Kakeya — DRAFTED

- **Under review:** finite-field reader + body-resistance framing + tiny `F_q^2`
  workbench. **No Kakeya result.**
- **Load-bearing call (Q3 ★):** is the **not-regime-2 fence** correct and strong
  enough — is "exact-maximal body-resistance pole" a sound reader framing, or an
  overclaim? (Their position: the direction-shadow is control-sufficient only
  trivially, because it *is* the direction.)
- **Owner-fill:** review page `kakeya.html` drafted (NOT PEER REVIEWED banner +
  `noindex`); the review URL + GitHub permalinks are staged into the packet's
  "Review Surface — Links" block, and the email is staged. **Still owner-gated:**
  deploy `kakeya.html` unlinked → `sundog.cc/kakeya`; push HEAD so the commit
  permalinks resolve; attach workbench + gallery screenshots + share-card PNG; pick
  reviewer.
- **Packet:** [`kakeya/EXTERNAL_REVIEW_PACKET.md`](kakeya/EXTERNAL_REVIEW_PACKET.md),
  [`SUNDOG_V_KAKEYA.md`](SUNDOG_V_KAKEYA.md).
- **Gate:** public `kakeya.html` launch + any inbound links.

### Hodge — DRAFTED

- **Under review:** reading-only — body = algebraic cycle / shadow = rational
  `(p,p)` class; reader + Phase-2 bridge + Phase-3 gallery spec. **No Hodge
  result.**
- **Load-bearing call:** **(a)** a category error / mis-stated known case (forms
  vs harmonic reps vs `(p,p)` classes vs cycles; or claiming-known something
  open), or **(b)** vacuous — *only* standard exposition → file
  `HODGE-FRONT-A-VACUOUS`? (6 Qs: translation fidelity, register ladder / CE1–3,
  existence-pole placement, Faraday/AB bridge, roster G1–G5, known-case wording +
  fences.)
- **Owner-fill:** `[Name]` / `[link]` / `[Your name]` in
  [`hodge/EXTERNAL_REVIEW_EMAIL_DRAFT.md`](hodge/EXTERNAL_REVIEW_EMAIL_DRAFT.md).
- **Packet:** [`hodge/EXTERNAL_REVIEW_PACKET.md`](hodge/EXTERNAL_REVIEW_PACKET.md),
  [`SUNDOG_V_HODGE.md`](SUNDOG_V_HODGE.md).
- **Gate:** any public Hodge educational artifact.

### Yang–Mills — SENT 2026-06-01 (Dr. Biagio Lucini)

- **Status:** **sent to Dr. Biagio Lucini** (lattice gauge theory, Swansea — SU(N)
  deconfinement / Polyakov-loop work, an ideal fit for the finite-T Polyakov
  target) via **LinkedIn DM**, 2026-06-01; awaiting reply. Packet re-pointed to the
  informative-null story 2026-05-31.
- **Under review:** the **informative powered null** — on a powered (split-half
  ICC 0.965), disjoint (CV-R² −0.332) finite-T Polyakov order-parameter target,
  the small-loop gauge-invariant signature carries no rank-locality
  (primary@5 0.3042 vs CTRL_RAND 0.3292). Lane PAUSED.
- **Load-bearing call:** **(a)** is "powered" overstated / the `12²×4`, `N_t=4`
  cell mis-conditioned, or **(b)** is the null the **textbook** outcome (small
  local loops can't resolve the non-local Polyakov order parameter)? (7 Qs:
  powered-test soundness, abelian-boundary explanation, powered-target audit,
  β-slate, finite-T artifacts, probe-ladder completeness, anti-p-hunting.)
- **Note:** Q2 (the new centerpiece) rests on the Faraday Phase 7/8 abelian-bridge
  — `yang-mills/2026-05-31_faraday_abelian_bridge_note.md`.
- **Owner-fill:** pick reviewer + fill; send packet + informative-null synthesis +
  bridge note.
- **Packet:** [`yang-mills/EXTERNAL_REVIEW_PACKET.md`](yang-mills/EXTERNAL_REVIEW_PACKET.md),
  [`SUNDOG_V_YANG_MILLS.md`](SUNDOG_V_YANG_MILLS.md).
- **Gate:** public Yang-Mills generality-gallery card.

### Riemann — RESOLVED-IN-HOUSE (2026-05-30)

- **Was under review:** the bounded null across three lanes (Z2 reflection,
  explicit-formula parity, nonlinear S2). **All three framing questions answered
  in-house:** Q1 by a sine-kernel / GUE control (observed D at z = −0.85), Q2 by a
  parity-identity proof, Q3 by a 100k-ordinate window / unfolding sweep; the
  representation-sector question closed at desk.
- **Remaining:** external review no longer gates the framing. Only optional
  pointer value (is the prime-conjugate-selector reopen a known device? a
  citation?) plus the publish decision, which is separate and owner-owned.
- **Packet:** [`riemann/EXTERNAL_REVIEW_PACKET.md`](riemann/EXTERNAL_REVIEW_PACKET.md)
  (retained as a finished record), [`SUNDOG_V_RIEMANN.md`](SUNDOG_V_RIEMANN.md).
- **Action:** **none for framing** — off the "waiting" list.

## Pre-review synthesis feedback (NOT external review)

> **What this is.** A synthesis of self-run model passes (Hyperbolic/Qwen +
> Wolfram Alpha), 2026-06-02. **It is not a human domain expert and does not
> satisfy any lane's review gate.** Logged here so it becomes the *first*
> review-feedback note for comparison once a real reply lands, and so the
> portfolio has a baseline of what a strong LLM reader flagged. **Standing
> instruction: noted, not acted on.** No artifact is edited on the strength of
> these quotes; they convert to action only if a human reviewer echoes them.

| # | Quote (verbatim) | Lane it maps to | If actioned, touches | Status |
| --- | --- | --- | --- | --- |
| 1 | "Safe with edits — strengthen the body-resistance algorithmic-information disclaimers to prevent conflation with regime-2 sufficiency, but the core boundary discipline is mathematically sound." | NSE-C1 (+ portfolio body-resistance framing) | the body-resistance / regime-2 fence wording in the separation statement §7 and `CROSS_SUBSTRATE_NOTES.md` | noted, not acted on |
| 2 | "The v6a null is the expected/textbook outcome because local Wilson loops cannot resolve the non-local Polyakov order parameter; it confirms the apparatus reports 'no structure' correctly but is not surprising — say so in the synthesis." | Yang-Mills (this is exactly its load-bearing call (b)) | the informative-null synthesis framing in the YM packet | noted, not acted on |
| 3 | "NSE framing seems conservative / basically right." | NSE-C1 (the desired Q4 answer) | nothing — this is the target verdict, recorded as the synthetic baseline | noted, not acted on |

**Why hold rather than act:** quote 1 *recommends an edit* and quote 2 *recommends
a synthesis addition*. Acting on a synthetic reader's edits before the human
review would (a) contaminate the very framing the human is being asked to judge,
and (b) collapse the value of logging "what the LLM said vs what the expert said."
These are pre-registered here precisely so the human reply can be diffed against
them. The team also signalled these can seed our **own** first review-feedback
note — i.e. when we send, we may *disclose* "an LLM synthesis pass already
suggested strengthening the body-resistance disclaimer; do you agree?" as an
honest prompt, without having silently pre-applied it.

## Send log

Append one row per dispatch so the portfolio's review state stays visible.

| Date | Lane | Reviewer (role, not always name) | Channel | Outcome / next check |
| --- | --- | --- | --- | --- |
| 2026-06-01 | Yang–Mills | Dr. Biagio Lucini (lattice gauge theory, Swansea) | LinkedIn DM | Awaiting reply (no reply yet) |
| 2026-06-02 | Navier–Stokes C1 | Prof. Edriss Titi (data assimilation / determining modes, Texas A&M) | email | Sent (per Titi draft Sender Note; owner to confirm date/channel). Awaiting reply |
| _staged_ | Navier–Stokes C1 | Prof. Eric Olson (continuous data assimilation / determining modes, UNR) | email | Parallel first-wave draft `internal/outreach/NSE_C1_OLSON_EMAIL_DRAFT.md`; attach the Olson-geared `PDE_C1_REVIEW_BUNDLE_OLSON.pdf` (CDA-framed cover, byte-identical body), then log send date here |