We are delighted to invite you to a stakeholder workshop on Thursday 6th March, at The Alan Turing Institute to showcase our findings in the context of Prof Liakata's Turing AI fellowship.

This will be a joint event with presentations from Prof. Liakata's Turing AI fellowship team on "Creating time-sensitive sensors from heterogeneous user-generated content” - the fellowship is ending in March)and a workshop themed around mental health, a major use case for the RAi UK Keystone project "Addressing sociotechnical limitations of Large Language Models in medical and social contexts" (AdSoLve).

These are Prof. Liakata's major projects. The Turing AI Fellowship findings showcase will inform some aspects of the new Keystone RAi UK AdSoLve project and the research connections between both projects will be of interest to all our stakeholders.

The event will take place at The Alan Turing Institute from 1000 until 1630. (Possibility to follow part of the meeting online)

Agenda:

10:00-10:30 Welcome coffee

10:30-10:50 Overview of the Turing AI Fellowship and link with AdSoLve, Prof. Maria Liakata, QMUL

10:50-11:10 Identifying moments of change with TempoFormer, Talia Tseriotou, QMUL

11:10-11:30 Capturing timelines of interest and identifying changes with neural Hawkes processes, Anthony Hills, QMUL

11.30-11.50 Timeline summarisation and improving temporal reasoning with LLMs, Jiayu Song, QMUL

11.50-12.10 Framework and benchmark for reasoning limitations in LLMs, Mahmud Akhter, QMUL

12.10-12.30 Evaluation of synthetic language data and LLM output, Jenny Chim, QMUL

12.30-12.45 Dashboard for mental health monitoring, Sebastian Lobbers, QMUL

12:45-13:30 Lunch

13.30-14.00 Turing AI Fellowship Partner talks with Dr Becky Inkster, Cambridge University and Prof. Dana Slonim, Bar-Ilan University.

14.00-15.00 AdSoLve Mental Health Use Case: Partner talks and Demo, Prof Domenico Giacco, University of Warwick; Jennifer Martin MindTech; Onno Kampman (MOH Ministry for Healthcare Transformation, Singapore)

15.00-16.30 Introduction to Workshop: Summarisation for therapy sessions; Dialogue for self-management – Discussion

16.30 End of the event

We would like to encourage as many to attend in person as it facilitates quality interactions and conversations. To attend in person, please reserve a spot now.

For Online participation:

We also understand many interested colleagues cannot attend the event in-person, but still wish to participate and be part of the conversation. To attend online, please leave your details HERE.

Please note that this event will involve data collection through audio and visual recordings, note-taking, and the collection of materials developed online or in person for use within AdSoLve project. By attending, you acknowledge and accept that.

Turing AI Fellowship Showcase & AdSoLve Co-Creation Workshop