Salvato in:
| Autori principali: | , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | https://arxiv.org/abs/2411.03651 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866916470145941504 |
|---|---|
| author | Alamdari, Parand A. Ebadian, Soroush Procaccia, Ariel D. |
| author_facet | Alamdari, Parand A. Ebadian, Soroush Procaccia, Ariel D. |
| contents | We consider the challenge of AI value alignment with multiple individuals that have different reward functions and optimal policies in an underlying Markov decision process. We formalize this problem as one of policy aggregation, where the goal is to identify a desirable collective policy. We argue that an approach informed by social choice theory is especially suitable. Our key insight is that social choice methods can be reinterpreted by identifying ordinal preferences with volumes of subsets of the state-action occupancy polytope. Building on this insight, we demonstrate that a variety of methods--including approval voting, Borda count, the proportional veto core, and quantile fairness--can be practically applied to policy aggregation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2411_03651 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Policy Aggregation Alamdari, Parand A. Ebadian, Soroush Procaccia, Ariel D. Artificial Intelligence Computer Science and Game Theory Machine Learning We consider the challenge of AI value alignment with multiple individuals that have different reward functions and optimal policies in an underlying Markov decision process. We formalize this problem as one of policy aggregation, where the goal is to identify a desirable collective policy. We argue that an approach informed by social choice theory is especially suitable. Our key insight is that social choice methods can be reinterpreted by identifying ordinal preferences with volumes of subsets of the state-action occupancy polytope. Building on this insight, we demonstrate that a variety of methods--including approval voting, Borda count, the proportional veto core, and quantile fairness--can be practically applied to policy aggregation. |
| title | Policy Aggregation |
| topic | Artificial Intelligence Computer Science and Game Theory Machine Learning |
| url | https://arxiv.org/abs/2411.03651 |