Postdoctoral Positions in Computational Neuroscience & AI

Postdoctoral research position at EPFL's A. Mathis Group focusing on computational neuroscience, AI-driven behavior analysis, and embodied agent training. The role involves developing open-source tools, advancing multimodal AI models, and mentoring students in a collaborative, open-science environment.

This postdoctoral role sits within the A. Mathis Group for Computational Neuroscience and AI (CNAI) at EPFL, directed by Professor Alexander Mathis. The position centers on using artificial intelligence to measure and interpret behavior, exploring how brains and embodied agents learn to control physical bodies, and investigating how the brain constructs a sense of its own body through proprioception. The group is known for creating widely adopted open-source software, training biomechanically realistic agents, and building AI models of sensorimotor processing.

Responsibilities

• Lead independent, innovative research projects that align with the lab's goals and the researcher's expertise.
• Contribute to the development and maintenance of open-source software, datasets, and evaluation benchmarks used by the research community.
• Provide guidance and mentorship to PhD and Master's students while fostering a collaborative team environment.
• Collaborate with internal and external partners, including neuroscience labs, ecological research groups, and clinical teams.
• Share research findings at international conferences and workshops.
• Create and manage behavioral datasets for various animal species.
• Design and implement methods for reconstructing animals and their environments.
• Develop techniques for understanding behavior from multi-view video and audio recordings.
• Improve multimodal large language models for detailed action understanding and reasoning.
• Help evolve DeepLabCut and other open-source tools used globally by research labs.
• Train muscle-actuated, biomechanically realistic agents to perform complex motor tasks using reinforcement learning, imitation learning, and curriculum-based methods.
• Create new approaches for full-body musculoskeletal control.

Requirements

• A completed or nearly completed PhD in computer science, applied mathematics, computational neuroscience, robotics, physics, or a closely related discipline.
• A strong publication record in relevant academic venues.
• Proficient programming skills in Python and experience with modern deep-learning frameworks.
• Excellent written and spoken English communication abilities.
• Self-motivated, collaborative, and dedicated to open science and reproducible research practices.
• Proven ability to work independently and effectively within a team.

Nice to have

• Background or familiarity with neuroscience or ethology.
• Experience or knowledge in motor neuroscience and biomechanics.

What the company offers

• A dynamic, collaborative research setting at the intersection of AI and neuroscience.
• Access to EPFL's high-performance computing infrastructure and the broader EPFL research ecosystem.
• A competitive salary aligned with EPFL postdoctoral standards.
• Excellent working conditions.
• Opportunities for international collaboration, conference travel, and professional development.

About the company

EPFL is a leading research university known for its focus on science and technology. The A. Mathis Group for Computational Neuroscience and AI (CNAI) is committed to open science and reproducible research, fostering a collaborative environment where researchers can advance the understanding of behavior, learning, and sensorimotor processing through AI.

• The team is dedicated to open science and reproducible research practices.

How to apply

Documents to submit:

• Motivation letter
• CV with publication list
• Two relevant publications
• Referee email addresses