| paper | d4-conf-atal-2018-0032 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | no |
The supplied paper has no numbered Theorem, Lemma, Corollary, or Proposition stating an algorithmic, complexity, or hardness result, so it provides no qualifying computational anchor. The proposed resident-type model adds collective utility aggregation to a paper about eliciting one user's utility, and cannot supply the missing result.
fails bit a — no named computational result to mirror
The proposed sum of utilities requires comparable cardinal scales across types, while the paper elicits ordinal comparisons and supplies no cross-user scale model.
fatal: True
No named computational result is mirrored; the paper's elicitation methods, experiments, and single-user traffic-policy application remain without a continuous-population analogue anchored in a result of this paper.
I can’t make an anchor-based positive case for this paper: the supplied text contains no numbered theorem, lemma, corollary, or proposition establishing a computational complexity result. Its named contributions are methods and empirical findings about query strategies, Gaussian-process preference models, and experiments. Those findings cannot honestly serve as the required complexity anchors.
The best adjacent idea is a future multi-user version of the Amsterdam traffic-policy study. Let each type be a complete preference and response model over the candidate policies, and let \(\mu_t\) be the fraction of a large group of residents with that type. An elicitation procedure could query comparisons or rankings from type groups, then select a policy to maximize a specified aggregate objective, such as \(\sum_t \mu_t u_t(v)\). A city with many residents and relatively few preference types is a plausible regime. The paper’s §8 explicitly lists multi-user elicitation as future work, which lends this extension some plausibility.
But that would change the paper’s central problem: it studies eliciting one specific user’s utility and choosing that user’s preferred policy. The proposed population model adds aggregation across users, and its objective would need to be specified. The paper’s continuous Pareto set and Gaussian-process domain concern policy outcomes and utility functions; they do not make the society continuous. Its user study and two-expert application also do not establish a high-multiplicity population regime.
So the honest positive foothold is a plausible research extension, not a continuous mirror of a named result in this paper. There is no anchor here for which to state the requested prover-ready question or predict whether its hardness transfers.
The strongest negative is that there is no computational complexity result in this paper to anchor a mirror. Its contributions are elicitation methods and empirical comparisons; it does not prove the complexity of a computational problem. The Gaussian process ranges over policy-value vectors, and the Pareto set may be infinite, but that continuizes the outcome space, not the society. Those features do not supply a population-complexity result.
The best foothold is the multi-user traffic-policy extension. One could let \(\mu_t\) be the share of residents with a complete preference and response type, query representatives of those types, and choose a policy for the group. That is plausible: the authors mention multi-user elicitation as future work, and their Amsterdam application concerns city policy. But making it computational requires choices the paper does not make: what aggregate objective to optimize, how to compare users’ utilities, and how to price or allocate queries across types. Pairwise comparisons give ordinal information, not the common cardinal scale needed for a summed-utility objective. A majority or ranking-based objective avoids that scale issue but defines a collective elicitation problem rather than the paper’s task of finding the policy preferred by one specific user.
So the stronger version of the proposed mirror is coherent, and it may be a worthwhile research direction. Its weakness as a *mirror of this paper* is that it changes the decision maker and adds the aggregation problem; there is no named computational result here whose continuous version it would be. I would make that negative case strongly against treating the paper as an anchor. The broader claim that no scenario could make a continuous population version worthwhile is less secure: the authors’ stated interest in multi-user elicitation and the city-policy setting make that extension plausible, even though it would be new work rather than a continuization of a result in this paper.
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.