Informations générales
Intitulé de l'offre : Data scientist (NLP) (M/F) (H/F)
Référence : UMR9194-ETIOLL-002
Nombre de Postes : 1
Lieu de travail : PALAISEAU
Date de publication : vendredi 28 mars 2025
Type de contrat : IT en contrat CDD
Durée du contrat : 7 mois
Date d'embauche prévue : 2 juin 2025
Quotité de travail : Complet
Rémunération : Between 2932,84 and 3557,58
Niveau d'études souhaité : BAC+3/4
Expérience souhaitée : 1 à 4 années
BAP : D - Sciences Humaines et Sociales
Emploi type : Ingénieur-e d’études en production, traitement, analyse de données et enquêtes
Missions
Fine tune and use various types of Large language models. Assess their performances. Use them on datasets to extract relevant information.
Activités
Train, fine tune LLMs
Create a reproducible pipeline, document the data
Read the relevant literature and keep abreast of ongoing transformations
Compétences
A precise knowledge of LLM and their environment for NLP applied to social sciences is required. Very good coding skills (Python) is a must. Knowledge of LLM architectures (Transformers)
Contexte de travail
The Center for Research in Economics and Statistics (CREST) is a CNRS scientific research unit specializing in the use of quantitative methods in the social sciences. The Computational Social Sciences (CSS) team at CREST focuses on digital data and artificial intelligence and their uses in the social sciences. The CSS team consists of approximately ten people, including five permanent researchers, a research engineer, and doctoral and postdoctoral students.
CREST is a quantitative social sciences research unit with approximately 100 researchers and 200 doctoral students.
Partial remote work is possible, with some travel to Paris.