Javier Cembrano

AAMAS Conference 2026 Conference Paper

Metric Distortion in Peer Selection

• Javier Cembrano
• Golnoosh Shahkarami

In the metric distortion problem, a set of voters and candidates lies in a common metric space, and a committee of 𝑘 candidates must be elected. The objective is to minimize a social cost, defined as a function of the distances between voters and their chosen representatives, while the voting rule only has access to ordinal preferences. The distortion of a rule is the worst-case ratio between the social cost of its outcome and that of the optimal committee, taken over all consistent preferences and metrics. Weinitiatethestudyofmetricdistortioninpeerselection, where voters and candidates coincide. We consider four objectives, obtained by combining two aggregation rules with two types of social cost. Under additive aggregation, an individual’s cost is the sum of their distances to all committee members; under 𝑞-cost, it is their distance to the 𝑞th closest member. The overall social cost is either utilitarian, given by the sum of all individual costs, or egalitarian, given by the maximum individual cost. Surprisingly, we find that even on the line metric, peer selection retains much of the hardness of the general case: Lower bounds remain strictly larger than one for all objectives, and cases where bounded distortion is impossible in general remain so here as well. On a positive note, cases with bounded distortion in the general setting achieve better constants in peer selection. For utilitarian cost, selecting the 𝑘 middle agents achieves a distortion between 1 and 2 under additive aggregation. Under 𝑞-cost, we show positive results for 𝑞 = 𝑘 = 2, but impossibility results largely carry over. For egalitarian cost, selecting the extremes yields an optimal distortion of 2 under additive aggregation and for 𝑞-cost with 𝑞 > 𝑘/3, while no bounded distortion is possiblewhen𝑞 ≤ 𝑘/3. Overall, ourresultsshowthatpeerselection on the line metric admits better constants than the general case, yet fundamental hardness barriers persist.

NeurIPS Conference 2025 Conference Paper

Impartial Selection with Predictions

• Javier Cembrano
• Felix Fischer
• Max Klimm

We study the selection of agents based on mutual nominations, a theoretical problem with many applications from committee selection to AI alignment. As agents both select and are selected, they may be incentivized to misrepresent their true opinion about the eligibility of others to influence their own chances of selection. Impartial mechanisms circumvent this issue by guaranteeing that the selection of an agent is independent of the nominations cast by that agent. Previous research has established strong bounds on the performance of impartial mechanisms, measured by their ability to approximate the number of nominations for the most highly nominated agents. We study to what extent the performance of impartial mechanisms can be improved if they are given a prediction of a set of agents receiving a maximum number of nominations. Specifically, we provide bounds on the consistency and robustness of such mechanisms, where consistency measures the performance of the mechanisms when the prediction is correct and robustness its performance when the prediction is incorrect. For the general setting where up to $k$ agents are to be selected and agents nominate any number of other agents, we give a mechanism with consistency $1-O\big(\frac{1}{k}\big)$ and robustness $1-\frac{1}{e}-O\big(\frac{1}{k}\big)$. For the special case of selecting a single agent based on a single nomination per agent, we prove that $1$-consistency can be achieved while guaranteeing $\frac{1}{2}$-robustness. A close comparison with previous results shows that (asymptotically) optimal consistency can be achieved with little to no sacrifice in terms of robustness.

SODA Conference 2025 Conference Paper

New Combinatorial Insights for Monotone Apportionment

• Javier Cembrano
• José Correa 0001
• Ulrike Schmidt-Kraepelin
• Alexandros Tsigonias-Dimitriadis
• Victor Verdugo