| paper | d4-conf-aaai-2020-1499 |
| 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 has no numbered theorem, lemma, corollary, or proposition asserting a computational result. Its polynomial-time and query-count claim is in prose, and its MTSP results are empirical, so no qualifying result anchors a continuous-population mirror.
fails bit a — no named computational result to mirror
No result is covered by a mirror; the prose complexity claim and numerical MTSP experiments do not qualify as anchors.
I can’t make the requested anchored case: the paper has no numbered Theorem, Lemma, Corollary, or Proposition asserting a complexity result. It does have Definition 1, Algorithm 1, computational claims in prose, and numerical MTSP results, so that is not the same as saying it has no computational content.
The nearest candidate is the paper’s claim that RIGA runs in polynomial time and uses polynomially many queries when the known-preference solver is polynomial and the population size \(S\) and number of generations \(M\) are polynomially bounded. Its argument gives at most \(S^2\) comparison queries per generation. But this is not a numbered result, and it concerns one decision maker with unknown preference parameters, not a society represented by a distribution over voter types. I therefore can’t use it as an anchor or give it a continuous problem and expected complexity class under your rules.
A possible population setting would be residents sharing finitely many preference-weight types over public projects, with mass representing each type’s share. Yet aggregating those residents’ preferences changes the paper’s single-decision-maker problem into a collective decision problem. Without a qualifying result to anchor that change, I don’t think it makes an honest positive case.
The proponent raised no qualifying anchor. The nearest candidate is RIGA’s polynomial-time and query-count claim, but the paper gives it as a conditional argument in prose, not as a numbered complexity result. The paper’s formal material includes a definition and an algorithm; its MTSP results are empirical. That is a weak foundation for a paper-anchored continuization.
Even granting that claim as an anchor, RIGA concerns one decision maker. The same person answers every comparison, and each answer narrows one set of possible preference parameters, \(\Omega_\Theta\). The algorithm’s “population” is a set of candidate solutions, not a population of people. And \(\Omega_\Theta\), though continuous, represents uncertainty about one person’s preferences; it is not a continuous society.
The proposed rescue—many residents with different weight types choosing a common project—has to supply a new collective decision rule. Under utilitarian aggregation of linear utilities, the types collapse to their average weight vector \(\bar\omega\), leaving RIGA with one aggregate preference. A different rule, such as majority choice or worst-case regret across types, changes the objective and the query oracle. Giving each person a separate decision instead produces independent runs. These may be reasonable models, but none carries over the paper’s single-decision-maker result unchanged.
A more ambitious rescue would make the MOCO instance a high-multiplicity allocation problem: residents are typed demands, a planner chooses an allocation, and the planner’s preferences over objectives are unknown. That is a plausible adjacent application. But it adds the population-dependent feasible set, the planner’s role, and a query protocol; the paper supplies no result about that model. Its general algorithm might be useful there, but that is a new application to formulate and analyze, not a continuous mirror established by this paper.
So the strongest negative case is against a direct, paper-anchored mirror: the paper has no numbered complexity result, and its central elicitation problem has no society whose mass can be continuized. I would not defend the universal claim that no scenario could make a worthwhile mirror. The general MOCO framework leaves plausible population-allocation applications open; ruling them out would go beyond what this paper supports.
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.