English
Catégorie
Recherche
Bandeau image
Croissant LLM
Chapo
A language model (LLM: Large Language Model) has been developed by the MICS laboratory and Illuin Technology.
Contenu
Named "Croissant LLM", the main features of this model are:- Sovereign: trained on the Jean Zay calculator with open data- Accountable: fully sourced , Read the paper (ARXIV)
English
Catégorie
Recherche
Bandeau image
Tower
Chapo
Focus on the launch of Tower, a multilingual 7B parameter large language model (LLM) optimized for translation-related tasks, developed in part by Nuno Guerreiro and Pierre Colombo at CentraleSupélec's MICS laboratory.
Contenu
We are thrilled to announce the release of Tower, a multilingual 7B parameter large language model (LLM) optimized for translation-related tasks. Towe, Learn more
English
Catégorie
Recherche
Bandeau image
NeuroLang, a language to better question brain activity
Chapo
Developed at the Inria center in Saclay, NeuroLang is a query language specific to the neurosciences. It helps researchers and clinicians to formalize their questions in order to analyze functional neuroimaging data more efficiently. These data often require a highly mathematical formulation before revealing their secrets.
Crédits bandeau
©MIND - Read figure legend at bottom of article
Contenu
86 billion neurons. 150,000 billion synapses. 180,000 kilometers of nerve fibers. And information circulating at over 400 kilometers/hour. The human b
MIND develops methods for exploiting neuroimaging data, Created in September 2023, MIND is a joint Inria project-team with CEA and Université Paris-Saclay. It is the partial successor to the former Parietal
English
Catégorie
Recherche
Bandeau image
CDS@DATAIA Challenges
Chapo
For several years, and in particular since July 2021 within the DATAIA institute, the Paris-Saclay Center for Data Science has been organizing data science challenges for students and researchers on the Saclay plateau. Here is a look at the latest collaborations.
Contenu
Implemented by the Paris-Saclay Center for Data Science (CDS) and co-organized with the DATAIA Institute, these various machine learning challenges ar
Challenge 1 | Prediction of the isotopic inventory in a nuclear reactor core, This challenge, organized in August 2021, was carried out with the support of DATAIA Institute, in collaboration with the Institut de Radioprotection
Challenge 2 | Detection and classification of ovarian follicles,       This challenge was realized with the support of DATAIA Institute, in collaboration with INRIA, CNRS, INSERM and INRAE.
Challenge 3 | Predict age from brain grey matter (regression), This challenge was realized with the support of DATAIA Institute, in collaboration with CEA NeuroSpin. This challenge gathered 31 participants and 33
Challenge 4 | Brain age regression with deep learning, This challenge was realized with the support of DATAIA Institute, in collaboration with CEA NeuroSpin. Edouard Duchesnay, Antoine Grigis (Universit
Challenge 5 | ATLAS Stroke Lesion Segmentation,     This challenge was realized with the support of DATAIA Institute, in collaboration with University of Southern California (USC).
Challenge 6 | Brain age prediction and debiasing with site-effect removal in MRI through representation learning, This challenge was realized with the support of DATAIA Institute, in collaboration with CEA NeuroSpin. Antoine Grigis, Benoît Dufumier, Edouard Duc
Challenge 7 | Bovine embryos survival prediction,   This challenge was realized with the support of DATAIA Institute, in collaboration with the Institut National de Recherche pour l'Ag
CDS@DATAIA, 0
English
Catégorie
Recherche
Bandeau image
Chapo
38 research papers from institutions of the DATAIA Institute's perimeter, have been accepted to the NeurIPS Conference.
Crédits bandeau
© NeurIPS
Contenu
NeurIPS (Neural Information Processing Systems), widely considered the world’s largest AI and machine learning research conference, has published the
Key figures
9454, papiers soumis
1900, papiers acceptés
38, font partie de l'écosystème DATAIA
English
Catégorie
Recherche
Bandeau image
Chapo
IFPEN and Inria's team TAU join forces to apply machine learning to computational fluid dynamics.
Contenu
The ML4CFD  project aims to enhance the performance of CFD simulations using  machine learning. Employing ML models in CFD simulations with complex, ML4CFD in detail
English
Catégorie
Recherche
Bandeau image
Chapo
The BioNLP Open Shared Tasks 2019 workshop presented 7 scientific tasks, including the Bacteria Biotope task, and the participants' results published by the Association for Computational Linguistics.
Contenu
The event, The BioNLP Open Shared Tasks 2019 workshop co-organized by INRA-MaIAGE and DBCLS (Tokyo) brought together in Hong Kong on November 4 the organizers an
The Bacteria Biotope task, The Bacteria Biotope task in its 4th edition, in addition to the names of microorganisms, habitats and geographical locations, includes phenotypes (e.
Robert Bossy présente la tâche Bacteria Biotope au BioNLP-OST 2019
Key figures
10, teams participated in the challenge
31, submissions
18, INRA experts involved
English
Catégorie
Recherche
Bandeau image
Chapo
The « PEPER » project: an architecture for collecting data on energy consumption per second at Télécom SudParis.
Contenu
Selected as part of the DATAIA Institute's « Calls for Research Projects » in 2018, the « PEPER » project aims to use neural networks for prediction a
Contenu textuel
The « PEPER » project, The world of electrical energy is facing significant structural changes: electricity use is constantly increasing and climate challenges require an in
<p><img alt="PEPER team" data-align="left" data-entity-type="file" data-entity-uuid="c0eb3c5d-db04-462b-89
English
Catégorie
Recherche
Bandeau image
Chapo
Challenges in extracting information from scientific texts to improve access to and use of scientific information.
Contenu
On 15 April 2019, the MaIAGE unit launched "Bacteria Biotope" (BB'19) and SeeDev'19 (Event extraction of genetic and molecular mechanisms involved in
English
Catégorie
Recherche
Bandeau image
Chapo
Produced with Campus Channel, ENSTA ParisTech offers a video to answer this question by providing its expertise.
Contenu
Today, Artificial Intelligence is everywhere and its progress concerns all sectors: transport, energy, complex systems, defence... It is therefore nec
Subscribe to Recherche