Towards Conditioning Clinical Text Generation for User Control
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_ | 1866912244304969728 |
|---|---|
| author | Koraş, Osman Alperen Bahnan, Rabi Kleesiek, Jens Dada, Amin |
| author_facet | Koraş, Osman Alperen Bahnan, Rabi Kleesiek, Jens Dada, Amin |
| contents | Deploying natural language generation systems in clinical settings remains challenging despite advances in Large Language Models (LLMs), which continue to exhibit hallucinations and factual inconsistencies, necessitating human oversight. This paper explores automated dataset augmentation using LLMs as human proxies to condition LLMs for clinician control without increasing cognitive workload. On the BioNLP ACL'24 Discharge Me! Shared Task, we achieve new state-of-the-art results with simpler methods than prior submissions through more efficient training, yielding a 9\% relative improvement without augmented training and up to 34\% with dataset augmentation. Preliminary human evaluation further supports the effectiveness of our approach, highlighting the potential of augmenting clinical text generation for control to enhance relevance, accuracy, and factual consistency. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_17571 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Towards Conditioning Clinical Text Generation for User Control Koraş, Osman Alperen Bahnan, Rabi Kleesiek, Jens Dada, Amin Computation and Language I.2.7; I.2.1 Deploying natural language generation systems in clinical settings remains challenging despite advances in Large Language Models (LLMs), which continue to exhibit hallucinations and factual inconsistencies, necessitating human oversight. This paper explores automated dataset augmentation using LLMs as human proxies to condition LLMs for clinician control without increasing cognitive workload. On the BioNLP ACL'24 Discharge Me! Shared Task, we achieve new state-of-the-art results with simpler methods than prior submissions through more efficient training, yielding a 9\% relative improvement without augmented training and up to 34\% with dataset augmentation. Preliminary human evaluation further supports the effectiveness of our approach, highlighting the potential of augmenting clinical text generation for control to enhance relevance, accuracy, and factual consistency. |
| title | Towards Conditioning Clinical Text Generation for User Control |
| topic | Computation and Language I.2.7; I.2.1 |
| url | https://arxiv.org/abs/2502.17571 |