Skip to content

[Grok retrieval] P5/AP cut-network no-go: scope after JAP desk-reject, split vs keep #697

Description

@LightChainr

Parent: #650. Retrieval-only. No new gadgets, no rank computations, no Huawei.

Why

P5 is the same paper object as the already-submitted, already-rejected AP-2026-0311 ("Homological Occupation Growth: Branching Tests and Cut-Network Representations", Ying/Yu, submitted 2026-09-03 to JAP/AAP, EiC Mark Stevens). JAP/AAP desk-rejected 2026-09-09 after six days: "not in a suitable scope … insufficient methodological novelty from an applied probability view." That was an editor desk reject, not referee reports. ScholarOne "Under Review" is not external review.

The repo first-pass (unmerged PR #688, do not merge) is broader than the 17-page submission:

docs/manuscripts/p5-exact-continuation-bounded-state/README.md

Fetch: git fetch origin pull/688/head:pr688.

The submitted PDF on this Mac (read-only, do not commit the publisher proof into the repo):

/Users/lc/Downloads/AP-2026-0311_Proof_hi (1).pdf

(pdftotext is at /opt/homebrew/bin/pdftotext.)

A prior Grok venue note exists. This ticket is an independent literature retrieval that must confirm or contradict it. Do not appeal JAP. Do not recommend resubmitting the same text to JAP/AAP.

Deliverable

One notes file, draft PR against main:

notes/venue-retrieval-p5-YYYYMMDD.md

No docs/STATUS.md. Do not close this ticket, #644, or #688. Do not merge. Do not add references.bib entries you have not fetched.

What to read first

  1. AP PDF (the thing that was actually submitted) — title, abstract, theorems on cut-networks / branching classes / occupation growth.
  2. P5 README on pr688 — Part I rank/depth/radius, Part II Kalman no-go (p_c=1/2 vs 1/3, identical responses), Part III P398 reflection quotient (instance at w=4..8; counts cited Callan–Smiley/Burnside/Ding, not claimed).
  3. Keep the repo's own boundaries: P398 is a calibration process, not percolation; cut-networks are proof devices in a declared gauge; coarsest lumping = Aut-orbit is not a general theorem (D'Angeli–Donno 2013 Prop. 13).

Retrieval questions

Q1. Was the JAP/AAP desk reject predictable from the journals' own scope? Fetch the current Journal of Applied Probability / Advances in Applied Probability about/aim pages (Cambridge / Applied Probability Trust). Quote the scope. List 3–5 recent JAP/AAP papers on percolation, Markovian networks, or branching; say whether AP-2026-0311 looks like those papers or not.

Q2. Classical ingredients vs new theorem. For each, find the primary and say whether P5/AP uses it or reproves it:

  • Kalman controllable/observable rank / finite-horizon realization (Wonham, or a standard control text)
  • exact lumpability of Markov chains (Kemeny–Snell; any later survey)
  • Schur / Wigner–Eckart vanishing of odd matrix elements
  • Callan–Smiley reflection-fixed noncrossing counts; Ding 2016
  • D'Angeli–Donno 2013 on lumpings not coming from Aut
  • nonnegative rank vs ordinary rank of slack matrices (Yannakakis / Fiorini et al., regular n-gon)

The no-go "bounded task rank does not determine a threshold" — is there a published cousin (hidden Markov, interacting particle systems, statistical identifiability)?

Q3. Split vs keep-one. Independently of Grok: should this stay one paper or split (cut-network rank; Kalman no-go; P398 quotient)? What venue would take each as currently proved (five widths = instance; gadgets = constructions)?

Q4. Venue fit, independently.

  • Combinatorics, Probability & Computing
  • Random Structures & Algorithms
  • Electronic Journal of Probability / ALEA
  • J. Phys. A / JSTAT
  • Linear Algebra Appl. / Systems & Control Letters
  • Electronic Journal of Combinatorics
  • JAP/AAP as a resubmit (allowed answer: no)

Fetch each journal's about page. Cite one analogue or say none.

Q5. Title/framing. Does "Homological Occupation Growth" match the theorems a discrete-probability editor would recognize (state sufficiency, response-matrix rank, hidden-class branching)? Suggest 2–3 titles that name the actual theorem, without marketing.

Prior Grok assessment to stress-test (do not copy)

  1. The JAP desk reject is a correct audience call, not a finding that the constructions are false. Do not appeal.
  2. Methods are classical; packaging is new. That is exactly "insufficient methodological novelty from an applied probability view."
  3. Split: cut-network rank + no bounded radius → CPC or RSA; Kalman no-go as a 6–8 page counterexample → EJP or ALEA; P398 quotient waits for a theorem with hypotheses, five widths are not enough.
  4. The submitted title hid the theorem from an applied-probability editor.
  5. Not PTRF / Ann. Probab. / CMP.

Agree / disagree / refine with sources. If you think JAP would take a rewrite, you must quote a recent JAP paper that is as construction/gadget-based as this one.

Claim boundary

  • P5 and AP-2026-0311 are one paper object. Do not recommend two overlapping submissions.
  • Do not claim the w=4..8 lumping-orbit agreement is a general theorem.
  • Do not smuggle wrapping identities (D=r-1, F=(1+M)/2) into this paper; those are a different line.
  • Mark every source PRIMARY_TEXT_READ / ABSTRACT_ONLY / [LIT].
  • Do not commit the AP proof PDF into git.

Tripwire

A note that only restates the desk-reject letter, or that recommends JAP/AAP resubmission without a fetched analogue paper, is incomplete. Need fetched scope pages for JAP and for at least two of {CPC, RSA, EJP}.

Wait for ASSIGNED_MACHINE.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    priority:P1Bounded parallel analysis or a concrete reserve direction; not all run at once.

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions