Franck Dernoncourt

ROAST enhances sentiment analysis by focusing on entire reviews.

This study investigates the effectiveness of FActScore in multiple languages.

A new dataset aims to create clearer summaries through user feedback.

New models enhance the identification of speakers in dialogue content.

Two specialized QA datasets aim to improve question-answering systems for Adobe Acrobat and Photoshop.

This study highlights a new method to fine-tune language models effectively.

ULLME enhances text embeddings and retrieval tasks using large language models.

LoCAL simplifies the way we interact with lengthy documents.

A new framework allows AI agents to create actions dynamically for better problem-solving.

Innovative system blends retrieval methods for accurate, specialized answers.

New methods tackle challenges of finding answers in visually rich documents.