Postdoctoral Researcher: Representation Learning and Multimodal Tokenization for Scientific Foundation Models. (M/F)
New
- Researcher in FTC
- 12 mounth
- Doctorate
This offer is available in English version
This offer is open to people with a document recognizing their status as a disabled worker.
Offer at a glance
The Unit
Groupe de recherche en Informatique, Image, Automatique et Instrumentation de Caen
Contract Type
Researcher in FTC
Working hHours
Full Time
Workplace
14032 CAEN
Contract Duration
12 mounth
Date of Hire
01/10/2026
Remuneration
€3,072 – €4,258 depending on experience
Apply Application Deadline : 14 May 2026 23:59
Job Description
Missions
The research is in the field of artificial intelligence for science (AI for Science). The primary mission is to overcome technical bottlenecks related to the representation of complex scientific data (graphs, 3D structures, spectra) within foundation models. The postdoctoral researcher will design innovative "tokenization" strategies to integrate these non-linguistic data types into Transformer-based architectures, ensuring the preservation of geometric and topological properties.
Activity
Conduct a literature review on representation learning (SSL, Geometric DL) applied to the sciences.
Develop and implement tokenization algorithms for molecular graphs, 3D structures, and spectral signals.
Train and evaluate multimodal foundation models on high-performance computing clusters (Jean Zay, CRIAN, LPC).
Collaborate with domain experts (physicists, chemists) to validate the scientific relevance of the learned representations.
Write scientific papers for major conferences and journals (NeurIPS, ICML, ICLR, or specialized physics/materials journals).
Present research findings at seminars and international conferences.
Your Profil
Skills
Education: PhD in Computer Science, Artificial Intelligence, Applied Mathematics, or Computational Physics/Chemistry with a strong ML focus.
Technical: Deep understanding of Deep Learning (Transformers, GNNs, Auto-encoders).
Programming: Proficiency in Python and PyTorch or TensorFlow/JAX frameworks.
Tools: Experience with GPU training and large-scale data management.
Soft Skills: Strong autonomy, intellectual curiosity, and ability to communicate effectively in an interdisciplinary environment.
Languages: Fluency in scientific English (written and spoken).
Your Work Environment
The project is part of a unique collaboration between three laboratories at the University of Caen / CNRS / ENSICAEN: GREYC (Computer Science), CRISMAT (Materials Science), and LPC (Subatomic Physics). The candidate will benefit from a stimulating interdisciplinary research environment and privileged access to local and national high-performance computing resources (Jean Zay). The position is based in Caen, a dynamic city located 2 hours from Paris.
Compensation and benefits
Compensation
€3,072 – €4,258 depending on experience
Annual leave and RTT
44 jours
Remote Working practice and compensation
Pratique et indemnisation du TT
Transport
Prise en charge à 75% du coût et forfait mobilité durable jusqu’à 300€
About the offer
| Offer reference | UMR6072-FREJUR0-014 |
|---|---|
| CN Section(s) / Research Area | Mathematics and mathematical interactions |
About the CNRS
The CNRS is a major player in fundamental research on a global scale. The CNRS is the only French organization active in all scientific fields. Its unique position as a multi-specialist allows it to bring together different disciplines to address the most important challenges of the contemporary world, in connection with the actors of change.
Create your alert
Don't miss any opportunity to find the job that's right for you. Register for free and receive new vacancies directly in your mailbox.