Metraipolis › archive › dc_b59033ef742e

Preference Elicitation for Policy Optimization and Application to Aligning Heart Transplantation with Human Values — arXiv:2608.28620

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.28620
License (as archived)
CC0-1.0
Archived
2026-10-17T07:41:45.700Z
SHA-256 of the archived text
307b60ccfceef94e9213a5c39bd10f76ec999f647deba2b52dd6c5d355a84391
Stable URL of this page
https://metraipolis.com/documento/dc_b59033ef742e

How to cite

Preference Elicitation for Policy Optimization and Application to Aligning Heart Transplantation with Human Values — arXiv:2608.28620 — arXiv (en), https://arxiv.org/abs/2608.28620 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_b59033ef742e (Metraipolis archive dc_b59033ef742e, sha256 307b60ccfceef94e9213a5c39bd10f76ec999f647deba2b52dd6c5d355a84391)
BibTeX
@misc{dcb59033ef742e,
  title = {Preference Elicitation for Policy Optimization and Application to Aligning Heart Transplantation with Human Values — arXiv:2608.28620},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.28620},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_b59033ef742e; sha256 307b60ccfceef94e9213a5c39bd10f76ec999f647deba2b52dd6c5d355a84391},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.28620. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Preference elicitation is essential for aligning AI systems with human values. Prior approaches (e.g., for organ allocation) often ask stakeholders to compare the decisions of an algorithm (e.g., patient A vs. patient B). Such a decision-level approach conflates the means with the ends. Instead, we elicit preferences directly over allocation outcomes to learn a utility function for policy optimization. We construct a novel preference elicitation algorithm for linear utilities that outperforms prior techniques in practice. Our algorithm has two phases. The first phase learns cutting planes through pairwise comparisons to rapidly shrink the space of possible attribute weights and warm-starts the second phase by eliminating dominated regions. The second phase then provably converges to the user's utility function. We apply our technique to heart transplant allocation, where a policy must balance competing objectives such as post-transplant outcomes, waitlist mortality, geographic ease, and equity. Using our algorithm, we conduct a user study to learn and aggregate a human-value-aligned utility function. We evaluate the linear representation against LLM-based methods supplied with users' natural language explanations, finding that none of these methods is more accurate than the learned linear utilities. We use the resulting utility function to optimize heart transplant policies that are significantly better aligned with human values. Compared to the hindsight optimum, the status quo policy achieves a competitive ratio of just 0.66, while our method is near-optimal with a competitive ratio of 0.95.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-01 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2608.28620 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-17T07:28:26.456Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).