Weronika Sieińska

Research project title: 
Large Language Models for Safe Human-Robot Interaction
Principal goal for project: 
To explore safety issues around Large Language Models for Human-Robot Interaction
Research project: 

I am involved in the EU H2020-ICT SPRING research project (Socially Pertinent Robots in Gerontological Healthcare). The objective of this project is to develop socially assistive robots with the capacity of performing multi-person interactions, and to validate the new technology based on the needs of gerontological healthcare. The robot used in the project is a social humanoid robot called ARI, which is developed by PAL Robotics. The conversational interface built in the robot is based on a Large Language Model (LLM), which is prompt engineered for the task of providing hospital patients and their companions with hospital-related information. In my research, I explore safety issues around LLMs for Human-Robot Interaction in the healthcare domain.

About me: 

I am a part-time Research Assistant at Heriot-Watt University involved in the EU H2020-ICT SPRING project. Formerly, I was an Assistant Engineer at Samsung R&D Institute Poland, where I worked on Samsung's voice assistant Bixby. I hold MSc and BEng degrees in Computer Science from Adam Mickiewicz University. I also studied for a year as an Erasmus+ exchange student at the University of Southern Denmark.

Research interests:

  • AI ethics and safety, AI alignment, responsible AI,
  • factuality, misinformation (fake news) detection and mitigation,
  • LLM safety, LLM hallucination, retrieval augmented generation, red-teaming.

Selected publications:

  • Multi-party Multimodal Conversations Between Patients, Their Companions, and a Social Robot in a Hospital Memory Clinic [link]
  • Human – Large Language Model Interaction: The dawn of a new era or the end of it all? [link]
  • A Multi-party Conversational Social Robot Using LLMs [link]
  • Multi-party Goal Tracking with LLMs: Comparing Pre-training, Fine-tuning, and Prompt Engineering [link]
  • Data Collection for Multi-party Task-based Dialogue in Social Robotics [link]
  • A Visually-Aware Conversational Robot Receptionist [link]
  • Developing a Social Conversational Robot for the Hospital Waiting Room [link]
  • Combining Visual and Social Dialogue for Human-Robot Interaction [link]
  • Coronabot: A Conversational AI System for Tackling Misinformation [link]

Research portals:
Google Scholar, Scopus, Research Gate, Semantic Scholar, ORCID, DBLP