3rd TAILOR conference
5-6 June, 2023 in Siena, Italy
The 3rd TAILOR conference will be a place to meet fellow AI scientist to discuss all aspects related to Trustworthy AI in general and TAILOR activities in particular. With a mix of plenary lectures, workshops and poster session, we hope everyone within the TAILOR network will be able to contribute.
Logistics
The conference venue La Certosa di Pontignano is located 8 km outside Siena. If you come by car, you can follow the driving instructions provided on the webpage of La Certosa.
Otherwise, connections to Siena from Pisa, from Florence and from Rome are good, both busses and trains run regularly, and are easily found by e.g. Google Maps.
For transport between Siena and the Certosa, we provide free shuttle buses. See Practicalities for more information.
The first shuttle will depart from Siena at 3 pm on Sunday 4 June. The next will depart from Siena at 6 pm on June 4. You can find the shuttle timetable on the Practicalities webapge.
Practicalities
You can find all the information for practicalities (bus shuttle, conference dinner) on the dedicated webpage.
Keynote speakers
Keynote speakers will be François Terrier, “Regulation, Standards, Methods and Tools for AI Safety.”, and Thomas G. Dietterich, “Competence Models For Machine Learning Systems“.
Poster session
Poster sessions will consist of two sessions, during which all posters will be displayed. At the start of each session, selected posters will be presented with a 1-minute pitch.
Here you can find the list of posters: https://tailor-network.eu/poster-session-3rd-tailor-conference/
Satellite meetings – June, 4th
Joint WP4-WP5 workshop
- 16:00-17:00 Invited talk by Hector Geffner
- 17:00-17:30 Quick presentations of the work in the two WPs
- 17:30-19:00 Brainstorming among WP4 and WP5 participants
WP7 Workshop
- 16.00-16.50 Task leaders will present the progress in their tasks and layout their future plans for the next half year.
- 17.00- 18.00 Discussion on the AutoAI roadmap. We’ll be discussing the current state and plans on improving.
We invite anyone interested to join the session!
Third Community Workshop 2023
The Third Community Workshop for ICT48 projects is organised by VISION on 7 June, at the University of Siena.
The workshop is designed specifically for the entire ICT-48 Community of AI4Media, ELISE, HumanE-AI-Net, TAILOR, VISION as well as representatives of the new NoEs, ELSA and euROBIN, in particular Project Coordinators, Scientific or strategic leadership, WP-leads as well as representatives of the European Commission.
Find the program and the registration modules on the event’s website: https://www.vision4ai.eu/community-workshop-2023/
Anyone that is going to the workshop is welcome to the TAILOR conference in the afternoon of June 6, and to join the social dinner in the evening!
If you want to join the afternoon session and the dinner, please contact Trine Vikinge (trine.vikinge@liu.se).
Hourly Schedule
Sunday 4 June
- 16:00 - 19:00
- Joint WP4-WP5 Workshop
- Details to follow
-
Speakers:
Giuseppe de Giacomo (UNIROMA1), Luc de Raedt (KU Leuven)
- 16:00 - 18:00
- WP7 Workshop
-
Speakers:
Annelot Bosman, Holger Hoos (ULEI)
- 18:00 - 19:00
- Registration
- 19:00 - 20:30
- Welcome reception at La Certosa
- 20:30
- Dinner at La Certosa
Monday 5 June
- 8:00 - 9:00
- Registration
- 9:00 - 10:00
- Welcome & TAILOR update
-
Speakers:
Fredrik Heintz
- 10:00 - 10:30
- Coffee break
- 10:30 - 12:30
- Poster session 1
- 12:30 - 14:00
- Lunch
- 14:00 - 15:30
- The TAILOR Handbook of Trustworthy AI
- Group work on fundamental properties of Trustworthy AI. Make sure to sign up for one of the groups, instructions and lists will be provided on-site.
-
Speakers:
Francesca Pratesi, Lena Tasse, Umberto Straccia (CNR)
- 15:30 - 16:00
- Coffee break
- 16:00 - 17:45
- Poster session 2
- 17:45 - 18:00
- Challenges
- Past challenges; New Challenge: Mind the Avatar's Mind; New Challenge: Crossword challenge
-
Speakers:
Marc Schoenauer, Marco Gori, Wico Mulder
- 18:00 - 18:30
- General assembly
- 20:00 - late
- Conference dinner
- Venue: La Certosa
Tuesday 6 June
- 9:00 - 9:05
- Introduction
-
Speakers:
Fredrik Heintz
- 9:05 - 10:00
- Keynote: Regulation, Standards, Methods and Tools for AI Safety
-
Speakers:
François Terrier
- 10:00 - 10:30
- Industry session
- 10:30 - 11:00
- Coffee break
- 11:00 - 12:30
- Industry session, continued
- 12:30 - 14:00
- Lunch
- 14:00 - 14:15
- Latest news from DG CNECT
- 14:15 - 15:30
- Road map and joint SRA
-
Speakers:
Fredrik Heintz, Michela Milano
- 15:30 - 16:00
- Coffee break
- 16:00 - 17:00
- Role of TAILOR in respect of LLM
- Panel discussione moderated by Fredrik Heintz; The purpose of the panel is to discuss the role of TAILOR with respect to LLMs in Europe. We will cover matters such as What are the main challenges for LLMs in Europe? and What are the opportunities for TAILOR wrt to LLMs?
-
Speakers:
Ana Paiva (IST), Holger Hoos (ULEI), Jasmin Schultz, Thomas G. Dietterich
- 17:00 - 18:00
- Keynote: Competence Models For Machine Learning Systems
- Every AI system, and certainly every machine learning system, should have an accurate model of its own competence. This talk will review our recent work on creating competence models for classifiers, computer vision systems, and reinforcement learning. We will discuss both input competence–is the system competent to handle a given query–and output competence–can the system provide calibrated statements about the quality of the answer it has produced? For input competence, our focus is on detecting queries that fall outside of the training data. We apply anomaly detection algorithms to address this task. For manually-engineered features, these work well, but we have discovered that features learned by deep neural networks are often inadequate for this task. The problem is that features are only learned for “directions of variation” that are present in the training data. An anomaly that varies in some way that was invariant in the training data will not be detected. For output competence, we have studied calibration (for classifiers) and trajectory-wise prediction intervals (for reinforcement learning). We will report progress on point-wise calibration, which is more challenging that the standard notion of set-wise calibration. Then we will describe our approach to creating trajectory-wise prediction intervals using the tools of conformal prediction. All of these results assume stationarity of the distribution that is generating the data. We will conclude by briefly discussing our efforts to extend this work to shifting distributions.
-
Speakers:
Thomas G. Dietterich
- 19:00
- Social event/Dinner
- On this last evening of the conference, we will fully enjoy the fantastic Tuscan countryside. Busses will take us from La Certosa to Brolio, located in the midst of the beautiful Chianti landscape. First, we will have a guided tour in the Castello di Brolio, and then dinner at the Osteria di Brolio. Brolio is part of the Ricasoli wine producing establishment. Transport back to both La Certosa and Siena is arranged. If you didn’t sign up for this yet, please get in touch with Trine.
Date
- 05 - 06 Jun 2023
Time
- 09:00 - 17:00
Location
- 5 Pontignano 53019, Castelnuovo Berardenga, Siena