Papers by Anna Shamray
An Evaluation of Classifiers for Mapping Generative LLM Responses to Answer Options of Multiple-choice Questionnaires (2026.eacl-srw)
Copied to clipboard
| Challenge: | Large language models are used to generate multiple-choice style questionnaires that were originally intended for humans in persona simulations. |
| Approach: | They investigate the performance of smaller LLMs for mapping LLM outputs into the available answer options of multiple-choice questionnaires. |
| Outcome: | The proposed model underperforms on three datasets with differing answer option complexity. |