| paper | d4-journals-geb-2014-0066 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | no |
The paper’s Theorem and Lemma characterize cost-sharing rules and their axiomatic properties; neither states a computational result. Aggregating agents with identical service constraints gives a plausible continuous allocation formula, but it extends the paper’s rules rather than mirroring a computational finding.
fails bit a — no named computational result to mirror
No result is covered by a continuous computational mirror. The characterization, lower-bound lemma, and examples concern allocation rules, not computational guarantees.
I can’t make an anchor-compliant positive case: the paper has no numbered result asserting a computational complexity claim such as NP-hardness, polynomial-time solvability, or FPT. Its unnumbered “Theorem” characterizes the usage-index family \(\phi_\pi\) by axioms; its unnumbered “Lemma” characterizes when the resulting rule satisfies the Unanimity Lower Bound. Neither is a computational result.
The nearest positive opening is that agents with the same service constraints \(D_i\) are already indistinguishable for the rule. One could give each service-constraint type a population mass and aggregate its cost share. But that would turn the paper’s axiomatic allocation formula into a population-distribution model; it would not mirror a named computational result of the paper.
The proponent’s central point is the strongest case against a mirror: the paper gives no computational result to mirror. Its Theorem characterizes usage indices by axioms, and its Lemma characterizes when an allocation rule satisfies a lower bound. Neither addresses the complexity of a computational problem. The examples apply those rules; they do not analyze algorithms.
The best rescue is to treat agents with the same service constraints \(D\) as one type. That is a natural high-multiplicity regime, and the paper’s anonymity and formulas make aggregation straightforward. For type mass \(\mu_D\), the share of item \(a\)’s cost assigned to that type under \(\psi_\pi\) would be proportional to
\[
\frac{\mu_D\left(\lvert D(a)\rvert/\lvert D\rvert\right)^\pi}
{\sum_E \mu_E\left(\lvert E(a)\rvert/\lvert E\rvert\right)^\pi}.
\]
But this extends an allocation formula; it does not continuize a computational finding in the paper.
One could add a computational question about evaluating the rule when service constraints have a compact representation, such as a graph. Yet the paper specifies no input encoding for those constraints. If minimal service sets are listed, the relevant counts can be computed from that list. If they are described compactly, the hard part may be counting minimal paths or other structures for an individual type; aggregating type masses then weights those per-type results. That could motivate a new algorithmic study, but it is not a mirror anchored in this paper’s results.
So the negative case is strong on the paper’s eligibility for this task: there are no anchors to defeat, and the proposed type aggregation supplies no missing computational theorem. It is weak as the universal claim that no worthwhile continuous computational question could arise. The paper gives a natural population model; it simply does not give the programme a computational result to continuize.
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.