Papers by Martial Pastor
Discourse Realization of Generics in Human and LLM-generated Texts (2026.acl-long)
Copied to clipboard
| Challenge: | Large Language Models produce texts that appear coherent and credible, even when their factual reliability is uncertain. |
| Approach: | They propose a text-level genericity score derived from clause-level annotations and apply it to argumentative essays produced by humans and LLMs. |
| Outcome: | The proposed model is less generic than LLM-produced arguments, the study shows . higher genericity correlates with less structured, paratactic structures, the research shows a. |
Enhancing Discourse Parsing for Local Structures from Social Media with LLM-Generated Data (2025.coling-main)
Copied to clipboard
| Challenge: | Existing discourse parsers do not generalize well across genres and text types. |
| Approach: | They propose to integrate large language models into RST discourse parsers to improve parser performance in a social media context. |
| Outcome: | The proposed model improves parser performance in a social media context without pre-identified discourse units. |