Advanced School in Applied Machine Learning: Uncertainty and Robustness of DL Models for Science

Administration

Description

An ICTP Meeting

We are excited to announce the Advanced School on Uncertainty and Robustness of Deep Learning (DL) Models for Science. This initiative, co-organized by ICTP alongside SISSA/TSDS and AREA, will take place at ICTP, in Trieste (Italy) from July 23rd to July 31st, 2026.

The school targets early-career researchers and PhD students, aiming to offer the participants practical tools to perform Uncertainty estimation (UQ) in DL models for scientific applications. There will be a special focus on scalability, sustainability, and advanced HPC methods for optimizing deep learning pipelines.

Lecturers from top-ranked research centers/universities and big tech companies (e.g. IBM, Nvidia) will cover the topics listed below:

Scientific DL
UQ in DL 4 Science
Simulation based Inference
Bayesian NN
Explainable DL for Science

Each topic will be complemented by the necessary tools to create efficient, scalable, and portable ML pipelines that exploit the capabilities of modern HPC infrastructures.

30 participants will be selected with attention to creating a cohort with balanced representation of countries (particularly, least developed countries), gender and under-represented groups.

This school will be the third activity in the field of advanced applied Machine Learning supported by the brand-new ICTP International Consortium for Scientific Computing (IComp).

Lecturers
J. Carnerero Cano, IBM Research, Ireland
D. Coscia, SISSA, Italy
C. De Nobili, Critiqality, Italy
J. Gonzalez Abad, Instituto de Física de Cantabria (IFCA), Spain
K. Karchev, University of Barcelona, Spain
F. Lanusse, CNRS, France
P. Mukhopadhyay, Cambridge, UK
M. E. Olmo, Barcelona Supercomputing Center, Spain
A. J. Scaffidi, SISSA, Italy
J. S. Shawe-Taylor, University College London, UK
H. Strobelt, IBM Research, USA

📢 The students are invited to follow the ceremony for the AI prize, that will be hosted at ICTP on July 23rd.

Research abstracts: Applicants are encouraged to submit an abstract for a poster presentation. During the application, please make sure to use our templates to format your abstract. Templates are available below for download.

📢 The following activity to be held from 20 to 23 July 2026 may also be of interest for you: Youth in High-Dimensions: Recent Progress in Machine Learning, High-Dimensional Statistics and Inference (smr4227)

Grants: A limited number of grants are available to support the attendance of selected participants, with priority given to participants from developing countries. There is no registration fee.

Material

Abstract Templates Hackaton award photo Poster Preliminary list of Participants Tutorial Materials group photo

Advanced School in Applied Machine Learning: Uncertainty and Robustness of DL Models for Science | (smr 4228)

Go to day

Thursday, 23 July 2026
- 14:30 - 18:00 The ICTP-IBM Brahmagupta AI Prize for Early Career Scientists Ceremony
```
The Ceremony will be livestreamed at: ictp.it/livestream
```
  Location: Leonardo Building - Budinich Lecture Hall
  - 14:30 Opening remarks 20'
    
    Speaker: Atish Dabholkar (Director, ICTP, Italy), Alessandro CURIONI (IBM, Switzerland)
  - 14:50 Announcement and motivation by the jury 10'
```
Remote
```
    Speaker: Marc MEZARD (Bocconi University, Italy)
  - 15:00 Award ceremony 10'
  - 15:10 Laudatio 20'
    
    Speaker: David S. WISHART (University of Alberta, Canada)
  - 15:30 Charting the unknown metabolome with biochemical AI 45'
    
    Speaker: Michael SKINNIDER (Princeton University, USA)
  - 16:15 Small, Specialised, Collaborative and Open-Weight: The Models Enabling Physical AI 45'
    
    Speaker: Eiman KANJO (NTU, UK)
  - 17:00 Group Photo 5'
  - 17:05 Refreshments 55' ( Leonardo Building Terrace )

Friday, 24 July 2026

08:45 - 09:00
- 08:45 Registration Formalities 15'

09:00 - 17:30

Location:

Adriatico Guest House - Kastler Lecture Hall

09:00 Opening Ceremony 30'

Speaker:

Alberto CAZZANIGA, (AREA Science Park, Italy), Serafina DI GIOIA (ICTP, Italy), Teodoro LAINO (IBM Research, Switzerland), Roberto TROTTA (SISSA, Italy)

09:30 Uncertainty-Aware Continual Learning for Adaptive Personalisation and Collaborative Intelligence 1h0'

Speaker:	Eiman KANJO (Nottingham Trent Univeristy, UK)
Material:	Video

10:30 Coffee Break 30'

11:00 Foundations of Bayesian Inference for Scientific Machine Learning 1h30'

Speaker:	Andrea SCAFFIDI (SISSA, Italy)
Material:	Video

12:30 Lunch Break 1h30'

14:00 Robust Foundation Models for Scientific Applications 1h30'

Speaker:	Francois LANUSSE (CNRS, France)
Material:	Video

15:30 Coffee Break 30'
16:00 Tutorial 1h30'

Speaker: Francois LANUSSE (CNRS, France)

Saturday, 25 July 2026

09:30 - 17:00

09:30 Efficient and Scalable Uncertainty Quantification in SciML - Part I 1h0'

Speaker:	Dario COSCIA (SISSA, Italy)
Material:	Video

10:30 Coffee Break 30'
11:00 Efficient and Scalable Uncertainty Quantification in SciML - Part II 1h30'

Speaker: Dario COSCIA (SISSA, Italy)
12:30 Lunch Break 1h30'
14:00 Free afternoon to explore the city 3h0'

Monday, 27 July 2026

09:30 - 17:30

09:30 Short talks by School Participants 1h0'
09:31 Automating License-Aware Full-Text Retrieval for Systematic Reviews: An End-To-End Scalable System to Reduce Reviewer Workload 20'

Speaker: Dmytro ZHUK (University of Vienna, Austria)
09:50 BRAHIM ABDELRAHMAN ELHASSAN NADA 20'

Speaker: IBRAHIM ABDELRAHMAN ELHASSAN NADA (University of Messina, Italy)
10:25 Group Photo (online Participants) 5'
10:30 Coffee Break 30'

11:00 Introduction to SBI for Scientific application 1h30'

Speaker:	Konstantin KARCHEV (University of Barcelona, Spain)
Material:	Video

12:30 Lunch Break 1h30'
14:00 Spiking Neural Network with Neuromodulated Attractors for Adaptive Robotic Quadruped Locomotion 20'

Speaker: Gabriel Gregorio TORRE DE SETA (University of Buenos Aires, Argentina)
14:20 Bayesian imaging inverse problems with scattering transforms 20'

Speaker: Sebastien Alexandre Nicolas PIERRE (École Normale Supérieure)
14:45 Tutorial - Part I 45'

Speaker: Konstantin KARCHEV (University of Barcelona, Spain)
15:30 Coffee Break 30'
16:00 Tutorial - Part II 1h30'

Speaker: Konstantin KARCHEV (University of Barcelona, Spain)

Tuesday, 28 July 2026

09:00 - 18:00

09:00 Multi-modal foundation model for physics (Walrus) 1h30'

Speaker:	Payel MUKHOPADHYAY (Cambridge, UK)
Material:	Video

10:30 Coffee Break 30'

11:00 AI Ethical Considerations 1h30'

Speaker:	John Stewart SHAWE-TAYLOR (University College London, UK)
Material:	Video

12:30 Lunch Break 1h30'

14:00 Uncertainty Quantification in Machine Learning Systems: Bayesian inference and beyond 1h30'

Speaker:	John Stewart SHAWE-TAYLOR (University College London, UK)
Material:	Video

15:30 Coffee Break 30'
16:00 Group Photo (in presence Participants) 5'
16:00 Poster session 2h0'

19:00 - 22:00
- 19:00 Social Aperitif with music 3h0' ( Pane Quotidiano, Barcola )

Wednesday, 29 July 2026

09:30 - 17:30

09:30 Tutorial 1h0'

Speaker: Payel MUKHOPADHYAY (Cambridge, UK)
10:30 Coffee Break 30'

11:00 Intro to AI for Climate, and Generative AI for climate downscaling 1h30'

Speaker:	Jose GONZALEZ ABAD (Instituto de Física de Cantabria, Spain)
Material:	Video

12:30 Lunch Break 1h15'

13:45 Lessons from Weather: Adapting GraphCast for Global Ocean Forecasting 1h45'

Speaker:	Stefano Campanella (OGS, Trieste)
Material:	Video

15:30 Coffee Break 30'

16:00 Climate DL Hackathon - Part II 1h30'

Speaker:

Jose GONZALEZ ABAD (Instituto de Física de Cantabria, Spain), Matias Ezequiel OLMO (Barcelona Supercomputing Center, Spain)

Thursday, 30 July 2026

09:00 - 17:30

09:00 Climate Hackathon winning Team announcement 30'

Speaker: Serafina Di Gioia (ICTP, Italy)

09:30 Intro to Explainable AI 1h0'

Speaker:	Hendrik STROBELT (IBM Research, USA)
Material:	Video

10:30 Coffee Break 30'
11:00 Tutorial 1h30'

Speaker: Hendrik STROBELT (IBM Research, USA)
12:30 Lunch Break 1h30'
14:00 Intelligent Agents and Physical Modeling of Transients in the Rubin Era 20'

Speaker: Phelipe Antonie DARC DE MATOS (Brazilian Center for Research in Physics, Brazil)
14:20 Uncertainty-Aware Vision-Language Models via Constrained Conformal Confidence Prediction 15'

Speaker: Andrea MELLIT (Mohamed 5 University of Rabat , Morocco)
14:35 Uncertainty estimation and robustness in deep learning models for retinal disease detection 15'

Speaker: Yousef EL MOURABIT (Sultan Moulay Slimane University, Morocco)

14:50 Deep Learning-Based Pipeline for Automated Fish Freshness Classification in Rastrelliger kanagurta (mackerel) and Sardinella longiceps 15'

Speaker:

Srishti BATRA (Permanent institute: National Centre for Biological Sciences Tata Institute of Fundamental Research, India)

15:05 Coffee Break 15'

15:20 Model Order Reduction meets Scientific Machine Learning for complex modelling: applications in parametric turbulent computational fluid dynamics 1h30'

Speaker:	Gianluigi ROZZA (SISSA, Italy)
Material:	Video

Friday, 31 July 2026

09:30 - 17:30

09:30 Factuality evaluation and Trustworthiness in LLM and AI agents for Science 1h30'

Speaker:	Javier CARNERERO CANO (IBM Research, USA)
Material:	Video

10:30 Coffee Break 30'
11:00 Tutorial 1h30'

Speaker: Javier CARNERERO CANO (IBM Research, USA)
12:30 Lunch Break 1h30'

14:00 UQ for reliable LLMs 1h30'

Speaker:	Andrea SANTILLI (NVIDIA, Italy)
Material:	Video

15:30 Coffee Break 30'

16:00 Collective properties of multi-agent systems 1h0'

Speaker:	Cristiano DE NOBILI (Critiqality, Italy)
Material:	Video

17:00 Closing Ceremony 30'

Speaker: Serafina DI GIOIA (ICTP, Italy)