Papers by Andrea Sensi

1 papers
Training Multi-Modal LLMs through Dialogue Planning for HRI (2025.findings-acl)

Copied to clipboard

Challenge: Existing approaches to enhance Multi-Modal Large Language Models (MLLMs) with explicit dialogue planning improves response accuracy and quality, and allows models trained in one language to transfer effectively to another.
Approach: They propose an approach that enhances Multi-Modal Large Language Models with a novel explicit dialogue planning phase that allows agents to refine their understanding of ambiguous commands.
Outcome: The proposed approach reduces hallucinations and improves task feasibility by fine-tuning and assessing Multi-Modal models in human-robot interaction scenarios.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations