25th and 26th of November
Sorbonne Cluster for Artificial Intelligence (SCAI)
NeurIPS@Paris 2025 is a 2-day event that takes place at Sorbonne Université on the 25th and 26th of November 2025. This event provides a local alternative to the NeurIPS conference. It aims to address some sustainability and inclusivity concerns in machine learning research. All students and researchers are welcome to attend.
Authors of all accepted papers at NeurIPS 2025 are allowed to present their work through an oral presentation and/or a poster session. We unfortunately cannot accommodate oral presentations for all accepted papers; a subset of accepted papers will be proposed for an oral presentation, based on scientific quality, thematic coherence, and diversity criteria.
Registration
The event is free, but registration is mandatory. Registration are now closed.
This event is in-person only: we will neither record presentations nor include online options.
Venue
This event will take place at Campus Jussieu (4 Place Jussieu, Paris 5ème).
To access by metro, go to Jussieu station on line 10.
The presentations will be held in amphithéâtre 25. All social events (poster sessions, mentor and career sessions, cocktail and lunch) will be held in SCAI and in the Esclangon building, which is located within Jussieu. A map of Jussieu is available here.
Schedule
The event is planned to start at 8:45am on November 25 and to end at 6:30pm on November 26.
It will feature oral presentations and poster sessions of NeurIPS 2025, as well as two opening keynotes on the 25th and 26th November (details below).
The schedule is available here. Two poster sessions are planned, with details available here.
Opening keynotes
This year, we have the honour to host Anne-Laure Ligozat and Gaël Varoquaux for the opening keynotes on the 25th November and 26th November, respectively.
-
Anne-Laure Ligozat will present a keynote on “Can we measure AI’s environmental cost?”. She is a Professor of computer science at LISN and ENSIEE. Ligozat’s research revolves around the environmental impact of digital technonology and AI in particular. She obtained her PhD from Paris-Sud University in computer science under the direction of Mich`ele Jardino.
-
Gaël Varoquaux will present a keynote on “Unpacking the AI progress”. Gaël Varoquaux is a research director working on data science at Inria (French computer science national research) where he leads the Soda team. He is also co-founder and scientific advisor of Probabl. Varoquaux’s research covers fundamentals of artificial intelligence, statistical learning, natural language processing, causal inference, as well as applications to health, with a current focus on public health and epidemiology. He also creates technology: he co-founded scikit-learn, one of the reference machine-learning toolboxes, and helped build various central tools for data analysis in Python. Varoquaux has worked at UC Berkeley, McGill, and university of Florence. He did a PhD in quantum physics supervised by Alain Aspect and is a graduate from Ecole Normale Superieure, Paris.
Scientific presentations
The presentations are divided into 6 thematic sessions, with details available here. Each presentation will last 10 minutes with 5 minutes for questions. In the interest of time, additional questions should be deferred to the poster session. Speakers will also be able to present their work in the corresponding poster session.
Day 1 : 25.11.2025
| Time | Description |
|---|---|
| 10:00 AM - 11:00 AM | Theory of Machine Learning |
| 11:30 AM - 12:30 PM | Machine Learning Methods and Algorithms |
| 02:00 PM - 03:00 PM | Deep Learning |
Day 2 : 26.11.2025
| Time | Description |
|---|---|
| 10:00 AM - 11:00 AM | Probabilistic Methods |
| 11:30 AM - 12:30 PM | Machine Learning Methods and Algorithms |
| 02:00 PM - 03:00 PM | Applications of Machine Learning and Statistics |
Mentoring
Participants to NeurIPS in Paris will have the possibility to take part in a mentoring session with senior researchers from academia and industry. The mentoring session will be at Sorbonne University'e from 5:30pm to 6:30pm on the 26th November. Details will be announced soon.
Sustainability ambitions
We aim to organize a local and sustainable alternative to in-person attendance to international ML conferences. To this aim:
- We strongly encourage participants to NeurIPS@Paris to travel to Paris by sustainable means.
- We encourage participants to consider as much as possible not flying to the main NeurIPS conference.
EurIPS
Extending the spirit of local and sustainable events, EurIPS proposes an officially endorsed event in Copenhagen between the 2nd and 7th December 2025, on the same period as the official conference and one week after NeurIPS@Paris. We encourage participants to consider, as much as possible, this ecological alternative.
Committees
Organizing committee
Julie Alberge (Inria)
Edwige Cyffers (ISTA)
Mathieu Even (Inria)
Maxime Haddouche (Inria)
Virginie Loison (Inria, Inserm)
Ambroise Odonnat (Noah’s Ark Lab, Inria)
Maxence Noble (Polytechnique)
Advising committee
Chloé-Agathe Azencott (Mines Paris-PSL)
Francis Bach (Inria)
Claire Boyer (Université Paris-Saclay)
Gérard Biau (Sorbonne Université)
Vianney Perchet (Criteo, ENSAE)
Jean-Philippe Vert (Owkin)
Sponsors
This year’s event is supported by Google DeepMind, Bioptimus, Cohere and G-Research.
Contact
If you have any questions, please contact us at neuripsinparis@gmail.com.