Complete Evidence Extraction with Model Ensembles: A Case Study on Medical Coding
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_ | 1866915994589462528 |
|---|---|
| author | Beckh, Katharina Heuser, Sven Rüping, Stefan |
| author_facet | Beckh, Katharina Heuser, Sven Rüping, Stefan |
| contents | High-stakes decisions informed by decision support systems require explicit evidence. While prior work focuses on short sufficient evidence, regulatory compliance and medical billing call for complete evidence: all relevant input tokens that support a decision. We formulate complete evidence extraction as a task and study it in a medical coding setting. Motivated by the Rashomon effect, we aggregate token-level evidence from multiple language models to increase evidence completeness. We perform a case study using existing equally-performing models, feature attributions, and a dataset with human-annotated evidence. Our results show that Rashomon ensembles significantly increase evidence recall while incurring only a small token overhead over individual models. Ensembles of only three models already outperform the best single model and recover information that individual models miss. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_07055 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Complete Evidence Extraction with Model Ensembles: A Case Study on Medical Coding Beckh, Katharina Heuser, Sven Rüping, Stefan Computation and Language Information Retrieval Machine Learning High-stakes decisions informed by decision support systems require explicit evidence. While prior work focuses on short sufficient evidence, regulatory compliance and medical billing call for complete evidence: all relevant input tokens that support a decision. We formulate complete evidence extraction as a task and study it in a medical coding setting. Motivated by the Rashomon effect, we aggregate token-level evidence from multiple language models to increase evidence completeness. We perform a case study using existing equally-performing models, feature attributions, and a dataset with human-annotated evidence. Our results show that Rashomon ensembles significantly increase evidence recall while incurring only a small token overhead over individual models. Ensembles of only three models already outperform the best single model and recover information that individual models miss. |
| title | Complete Evidence Extraction with Model Ensembles: A Case Study on Medical Coding |
| topic | Computation and Language Information Retrieval Machine Learning |
| url | https://arxiv.org/abs/2511.07055 |