Guardado en:
| Autores principales: | , , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | https://arxiv.org/abs/2603.17832 |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866914406748651520 |
|---|---|
| author | Hernandez, Jefferson Saha, Swarnadeep Whitehouse, Chenxi Parekh, Sanjeel Murdock, Calvin Li, Yuliang Brimijoin, W. Owen Ithapu, Vamsi Krishna Ananthabhotla, Ishwarya |
| author_facet | Hernandez, Jefferson Saha, Swarnadeep Whitehouse, Chenxi Parekh, Sanjeel Murdock, Calvin Li, Yuliang Brimijoin, W. Owen Ithapu, Vamsi Krishna Ananthabhotla, Ishwarya |
| contents | In this work, we probe the ability of a language model to demonstrate spatial reasoning from unstructured text, mimicking human capabilities and automating a process that benefits many downstream media applications. Concretely, we study the narrative-to-play task: inferring stage-play layouts (scenes, speaker positions, movements, and room types) from text that lacks explicit spatial, positional, or relational cues. We then introduce a dramaturgy-inspired deterministic evaluation suite and, finally, a training and inference recipe that combines rejection SFT using Best-of-N sampling with RL from verifiable rewards via GRPO. Experiments on a text-only corpus of classical English literature demonstrate improvements over vanilla models across multiple metrics (character attribution, spatial plausibility, and movement economy), as well as alignment with an LLM-as-a-judge and subjective human preferences. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_17832 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Text-to-Stage: Spatial Layouts from Long-form Narratives Hernandez, Jefferson Saha, Swarnadeep Whitehouse, Chenxi Parekh, Sanjeel Murdock, Calvin Li, Yuliang Brimijoin, W. Owen Ithapu, Vamsi Krishna Ananthabhotla, Ishwarya Computation and Language Artificial Intelligence Machine Learning In this work, we probe the ability of a language model to demonstrate spatial reasoning from unstructured text, mimicking human capabilities and automating a process that benefits many downstream media applications. Concretely, we study the narrative-to-play task: inferring stage-play layouts (scenes, speaker positions, movements, and room types) from text that lacks explicit spatial, positional, or relational cues. We then introduce a dramaturgy-inspired deterministic evaluation suite and, finally, a training and inference recipe that combines rejection SFT using Best-of-N sampling with RL from verifiable rewards via GRPO. Experiments on a text-only corpus of classical English literature demonstrate improvements over vanilla models across multiple metrics (character attribution, spatial plausibility, and movement economy), as well as alignment with an LLM-as-a-judge and subjective human preferences. |
| title | Text-to-Stage: Spatial Layouts from Long-form Narratives |
| topic | Computation and Language Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2603.17832 |