| paper | d4-conf-aaai-2022-0266 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | unclear |
Das states no numbered theorem, lemma, corollary, or proposition asserting an algorithm, complexity bound, or hardness result. The §5.3 allocation discussion motivates future work but leaves the objective and fairness constraints open, so the proposed mass-allocation problem would be a new formalization rather than a continuous analogue of a result in this paper.
fails bit a — no named computational result to mirror
The §5.3 discussion leaves the welfare objective and fairness constraints unspecified, so the allocation sketch is a new formalization rather than a mirror of a problem Das defines.
fatal: True
No result is mirrored. The §5.3 discussion motivates optimization with fairness constraints; the paper's other discussions of prediction, incentives, and local-justice principles also state no computational result by Das.
Das’s paper contains no numbered theorem, lemma, corollary, or proposition asserting a computational result. It is a perspective paper that frames allocation questions and calls for more work. In §5.3, the claim that exact solutions are “typically hard to find” motivates a research direction; it is not a formal hardness result. The formal results mentioned in the text belong to cited papers, so they cannot serve as anchors from this paper.
There is a plausible unanchored mirror in the paper’s discussion of allocating scarce resources under fairness constraints. For example, in a large batch allocation of homelessness services, a type could specify all decision-relevant features, predicted outcomes under each service, and fairness-group membership. Let \(\mu_t\) be the fraction of the population of type \(t\), and let \(x_{t,r}\) be the mass of type \(t\) assigned to service \(r\). One could optimize predicted population welfare subject to resource-capacity and policy fairness constraints. A regime with many applicants and relatively few recurring types is credible for such allocations.
That sketches a continuous-population research problem the author might recognize. But it does not mirror a named computational result by Das, so there is no honest anchor to restate, no per-anchor problem to hand to a prover, and no result-based complexity direction to predict.
The proponent names no result of Das’s to mirror, so there are no anchors to defeat individually. That is the strongest objection: this is a perspective paper, not a computational study. Its claim that exact solutions are “typically hard to find” is a motivation, not a defined problem or a complexity result. The formal results it mentions belong to cited work.
The strongest candidate is the §5.3 discussion of optimizing welfare subject to fairness constraints. A large batch allocation of shelter services, for example, could group applicants by all decision-relevant features, predicted outcomes, and group membership, then assign mass of each type to services under capacity and policy constraints. Repeated applicant categories make that a credible high-multiplicity regime. But Das supplies none of the formal choices needed to make this a particular computational problem: the allocation rules, fairness constraints, welfare inputs, and encoding are left open. With additive predicted welfare and linear constraints, one natural version is an LP. Adding exchange cycles, complementarities, strategic reports, or uncertain predictions could produce richer problems, but those additions would determine the algorithmic question; they are new formalizations, not mirrors of a problem Das defines.
That does not show such research would be uninteresting, and the plausible batch-allocation scenario makes a universal claim that *no* continuous mirror could be worthwhile hard to defend. The negative case is strong only in the narrower, paper-specific sense: Das offers no computational result whose continuous counterpart this programme can chart. The proponent’s proposal may be a worthwhile new direction inspired by the paper, but it is not an anchored mirror of one of its results.
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.