LLM-as-a-Judge: Toward World Models for Slate Recommendation Systems
Fuente:
arXiv
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866912691612811264 |
|---|---|
| author | Bonin, Baptiste Heuillet, Maxime Durand, Audrey |
| author_facet | Bonin, Baptiste Heuillet, Maxime Durand, Audrey |
| contents | Modeling user preferences across domains remains a key challenge in slate recommendation (i.e. recommending an ordered sequence of items) research. We investigate how Large Language Models (LLM) can effectively act as world models of user preferences through pairwise reasoning over slates. We conduct an empirical study involving several LLMs on three tasks spanning different datasets. Our results reveal relationships between task performance and properties of the preference function captured by LLMs, hinting towards areas for improvement and highlighting the potential of LLMs as world models in recommender systems. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_04541 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | LLM-as-a-Judge: Toward World Models for Slate Recommendation Systems Bonin, Baptiste Heuillet, Maxime Durand, Audrey Information Retrieval Artificial Intelligence Modeling user preferences across domains remains a key challenge in slate recommendation (i.e. recommending an ordered sequence of items) research. We investigate how Large Language Models (LLM) can effectively act as world models of user preferences through pairwise reasoning over slates. We conduct an empirical study involving several LLMs on three tasks spanning different datasets. Our results reveal relationships between task performance and properties of the preference function captured by LLMs, hinting towards areas for improvement and highlighting the potential of LLMs as world models in recommender systems. |
| title | LLM-as-a-Judge: Toward World Models for Slate Recommendation Systems |
| topic | Information Retrieval Artificial Intelligence |
| url | https://arxiv.org/abs/2511.04541 |