PhD in Multimodal Generative AI for Materials Design at EPFL (LTS2 Lab, Lausanne)
Job Description
EPFL’s LTS2 lab is hiring two students for a PhD in generative AI for materials design, building multimodal generative models, latent diffusion and reinforcement learning methods that connect to an automated platform for materials synthesis within a large ERC Synergy project in Lausanne.
At a glance
- Position: Two PhD students (EPFL reference 2048)
- Where: LTS2 Lab, EPFL, Lausanne, Switzerland
- Duration: Four years; one-year fixed-term contracts renewed up to four years
- Admission: You must also apply and be admitted to the EPFL doctoral programme in Electrical Engineering
- Apply: open until filled (re-checked 24 October 2026)
What the generative AI PhD in materials design involves
- Develop new multimodal generative AI architectures, latent diffusion processes and reinforcement learning methods for materials discovery
- Link the models to an automated platform that carries out experimental synthesis
- Bring in data from experimental characterisation and from simulations
- Work with materials scientists on de novo simulation, synthesis and characterisation
- Publish, present at international conferences and supervise student projects
Why this role matters
Generative models for materials often stop at a list of candidate structures. Here the model sits in a loop with robots that make and measure the materials, so its proposals are tested for real and the results feed back into training. The project also offers research stays at partner institutions in Belgium and France.
What EPFL is looking for
- An outstanding MSc in engineering, computer science, physics, applied mathematics or a related field
- A strong analytical background
- Skill in, or willingness to learn, generative deep learning: latent diffusion, multimodal models, statistics and learning theory
- Self-drive, problem-solving ability and interest in working across disciplines
- Professional command of written and spoken English
Pay and location
The ad offers four years with competitive pay but does not give a figure; EPFL publishes its doctoral assistant salary scale on its website. The post is in Lausanne and includes research internships at institutions in Belgium and France.
How to apply
Apply through the official EPFL job posting. EPFL’s posting does not give a closing date, so the role is open until filled; ResearchJobs.in will re-check this listing on 24 October 2026. Submit a motivation letter, a detailed CV, contact information and at least two referees through the posting, and apply separately to the EPFL doctoral programme in Electrical Engineering (EDEE). Check the posting for location and work-authorisation details before you apply.
See all our artificial intelligence jobs, or browse more research jobs on ResearchJobs.in.
Hiring institution: EPFL (École polytechnique fédérale de Lausanne)
Official advertisement: careers.epfl.ch
How to prepare for this application
- Apply to the EDEE doctoral programme in parallel; the lab cannot hire you without admission
- Show a generative model you have built or reproduced (a diffusion or flow model is ideal), with code and a short write-up
- Read recent work on generative models for crystals and molecules, and on closed-loop 'self-driving' labs, so you can discuss where they fail
- Explain how you would handle multimodal data such as structures, spectra and synthesis conditions in one model
- Name referees who can speak to your independent research, not only your grades
About EPFL
EPFL is the Swiss Federal Institute of Technology in Lausanne. The LTS2 signal processing lab works on machine learning and signal processing, with applications ranging from biology to neuroscience.
