Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
Benchmark Integrity and Reasoning-Trace Errors in Medical Question Answering With Large Language Models: Mixed Methods Study With Sparse Autoencoders
Saved in:
| Published in: | Journal of Medical Internet Research |
|---|---|
| Format: | Online Article RSS Article |
| Published: |
2026
|
| Subjects: | |
| Tags: |
No Tags, Be the first to tag this record!
|
| _version_ | 1868553926420201472 |
|---|---|
| collection | WordPress RSS FRELIP Feed Integration |
| container_title | Journal of Medical Internet Research |
| description | |
| discipline_display | Communications |
| discipline_facet | Communications |
| format | Online Article RSS Article |
| genre | Journal Article |
| id | rss_article:91270 |
| institution | FRELIP |
| journal_source_facet | Journal of Medical Internet Research |
| last_indexed | 2026-06-20T21:44:45.448Z |
| publishDate | 2026 |
| publishDateSort | 2026 |
| record_format | rss_article |
| spellingShingle | Benchmark Integrity and Reasoning-Trace Errors in Medical Question Answering With Large Language Models: Mixed Methods Study With Sparse Autoencoders Communications General Communications |
| sub_discipline_display | General |
| sub_discipline_facet | General |
| subject_display | Communications General Communications |
| subject_facet | Communications General Communications |
| title | Benchmark Integrity and Reasoning-Trace Errors in Medical Question Answering With Large Language Models: Mixed Methods Study With Sparse Autoencoders |
| title_alt | Integridad de los Puntos de Referencia y Errores de Traza de Razonamiento en la Respuesta a Preguntas Médicas con Modelos de Lenguaje Grande: Estudio de Métodos Mixtos con Autoencoders Dispersos Intégrité des références et erreurs de trace de raisonnement dans le question-réponse médical avec les grands modèles de langage : Étude de méthodes mixtes avec autoencodeurs parcimonieux Integridade de Benchmark e Erros de Rastreamento de Raciocínio em Respostas a Perguntas Médicas com Grandes Modelos de Linguagem: Estudo de Métodos Mistos com Autoencoders Esparsos |
| title_auth | Benchmark Integrity and Reasoning-Trace Errors in Medical Question Answering With Large Language Models: Mixed Methods Study With Sparse Autoencoders |
| title_es_txt | Integridad de los Puntos de Referencia y Errores de Traza de Razonamiento en la Respuesta a Preguntas Médicas con Modelos de Lenguaje Grande: Estudio de Métodos Mixtos con Autoencoders Dispersos |
| title_fr_txt | Intégrité des références et erreurs de trace de raisonnement dans le question-réponse médical avec les grands modèles de langage : Étude de méthodes mixtes avec autoencodeurs parcimonieux |
| title_full | Benchmark Integrity and Reasoning-Trace Errors in Medical Question Answering With Large Language Models: Mixed Methods Study With Sparse Autoencoders |
| title_fullStr | Benchmark Integrity and Reasoning-Trace Errors in Medical Question Answering With Large Language Models: Mixed Methods Study With Sparse Autoencoders |
| title_full_unstemmed | Benchmark Integrity and Reasoning-Trace Errors in Medical Question Answering With Large Language Models: Mixed Methods Study With Sparse Autoencoders |
| title_pt_txt | Integridade de Benchmark e Erros de Rastreamento de Raciocínio em Respostas a Perguntas Médicas com Grandes Modelos de Linguagem: Estudo de Métodos Mistos com Autoencoders Esparsos |
| title_short | Benchmark Integrity and Reasoning-Trace Errors in Medical Question Answering With Large Language Models: Mixed Methods Study With Sparse Autoencoders |
| title_sort | benchmark integrity and reasoning-trace errors in medical question answering with large language models: mixed methods study with sparse autoencoders |
| topic | Communications General Communications |
| url | https://www.jmir.org/2026/1/e90061 |