| paper | d4-conf-wine-2023-0010 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | unclear |
The paper reports simulation findings and uses an integer programming formulation, but contains no numbered theorem, lemma, corollary, or proposition stating a computational result. The opponent’s mass allocation model is a plausible computational follow-on, but it cannot supply the required result-level anchor, so bit (a) fails.
fails bit a — no named computational result to mirror
No named computational result is available to anchor a mirror; the paper's learning algorithms and simulated performance comparisons are not covered by a result-level continuous analogue.
There is no qualifying anchor in the paper. It contains no numbered Theorem, Lemma, Corollary, or Proposition asserting a computational complexity result.
The closest candidate is the statement in §1.1 that selecting a maximum weight collection of node-disjoint cycles and paths is NP-hard. It is unnumbered background, not a result proved by this paper. The paper’s own findings are simulation results reported in tables, not named complexity results. So under your anchor rule, there is no per-anchor continuous problem to state or classify.
The appended case is right about the formal gap: this paper proves no numbered complexity theorem, and its main findings are simulation results. But that gap does not establish the broader claim that no worthwhile continuous mirror exists. The paper explicitly identifies a computational core—maximum weight selection of node-disjoint exchange cycles and chains—and uses an integer program to solve it in every matching round. A mirror of that core would be a computational follow-on to the paper, even if it is not a theorem-level restatement.
Consider a national program with repeated cohorts of interchangeable patient–donor types. Define each type by all the features the allocation uses, including its compatibility profile and priority weight; include waiting-age strata if the objective tracks waiting time. Let \(\mu_t\) be the rational mass of type \(t\), and let \(Q\) be the feasible short cycles and altruist-started chains. A mass allocation can be written as
\[ \max_{x\ge 0}\ \sum_{q\in Q} w_q x_q \quad\text{subject to}\quad \sum_{q\in Q} b_{tq}x_q\le \mu_t\quad\text{for each type }t, \]
where \(b_{tq}\) counts how many participants of type \(t\) pattern \(q\) uses. Each transplant remains an integral bundle for each person; it is the mass of identical participants assigned to different feasible patterns that can split. For fixed cycle and chain caps \(C,P\), the type-labelled patterns can be enumerated in \(O(\tau^C+\tau^P)\) possibilities, giving an LP polynomial in the supported type count and input bit length when the caps are constants. If the caps are part of the input, pricing patterns becomes a search over long compatible chains or cycles, giving a natural boundary for the complexity study. The paper’s negative altruist weight also has a direct mass interpretation: it prices use of a scarce type of donor.
This model needs a credible high-multiplicity regime. In particular, cPRA intervals alone do not define complete types if actual crossmatch compatibility differs within an interval; compatibility must be part of the type, even if that increases \(\tau\). And if altruists remain a fixed small count as the population grows, their mass vanishes, so the mirror must use the positive per-capita altruist arrival regime the paper itself explores. Those are real constraints on the proposed scenario, not reasons that no such scenario can exist.
So the strongest case against “no worthwhile mirror, in any scenario” is a type-mass version of the paper’s exchange-selection problem, with a tractability boundary and a dynamic extension over cohort masses. Its weakness is also clear: the paper does not establish the computational result being mirrored, so under a strict requirement for a named theorem proved by the authors, this is an author-recognizable computational continuation rather than a faithful result-level mirror.
The adversarial triple: the proponent anchors on up to three named results; the opponent sees that case and must defeat every anchor; the judge decides which case convinced it. These are the pipeline’s own outputs, generated by tools/triple_run.py — no human edited them. The paper’s own text is not reproduced here beyond the quoted statement above.