Project description

Applications are invited for a Research Support Tech in the Theoretical and Cognitive Neuroscience (TCN), Moreno Bote Lab, https://www.upf.edu/web/tcn, led by Ruben Moreno Bote, Universitat Pompeu Fabra (Barcelona). The candidate will form part of the project Theory of maximally occupying multi-agents funded by the Air Force Defense Research Sciences Program, Office of Scientific Research United States Air Force. The position is for 6-12 months. The project aims to study the computational mechanisms underlying complex sensorimotor control in humans, other animals, and artificial agents. We will use stochastic optimal control theory to uncover general principles of motor control and to develop semi-analytical algorithms for problems that include realistic sources of sensorimotor variability, such as sensory, motor, and internal noise. The project has the objective of providing interpretable solutions to complex motor-control problems and clarifying how biological systems might handle multiple interacting noise sources; this would also allow us to develop inverse optimal control tools to explain human sensorimotor behavior under rationality principles. Finally, we aim to test whether these optimal strategies emerge in biologically constrained recurrent networks, thereby generating experimentally testable predictions about neural dynamics and sensorimotor computation. The optimal-control part of the project will study time-varying linear dynamical systems with quadratic costs and additive and multiplicative noise, extending the classical Linear-Quadratic-Gaussian framework to more realistic sensorimotor settings.

Using

Lagrange multiplier techniques, we aim to derive iterative semi-analytical algorithms for these stochastic control problems,



which can then be embedded in inverse optimal control pipelines to fit human behavioral data through maximum-likelihood approaches. In parallel, full-rank and low-rank recurrent neural networks will be trained on the same tasks using gradient-based optimization methods, such as Adam.

Tasks to be performed

Tasks are to assist with: - Analyzing behavioral data - Developing and testing algorithms and pipelines - Training neural networks. - Preparing publications The duration is for 6 – 12 months.

Group and complement: Group 3 + Level U

Dedication and working hours: part time (32,5h/week).

Planned remuneration approx: 27.838,56 € gross per year (32,5 hours/week)

Financing fund: This position is funded by the Air Force Office of Scientific Research (AFOSR) under award number FA9550-25-1-0363. PRUSA00525 - 4040-0001 MORENO BOTE, RUBEN

Requirements

● Batxillerat-FP2/ High school-FP2.

The expected start date is 14 September 2026 , the job is in Barcelona.

Selection criteria: The selection of the candidates will be made through evaluation of the curriculum and, where appropriate, with the carrying out a test and/or interview. Valuation will be as follows: The minimum score to pass the selection process is 70 points. The candidate with the highest score in the selection process will be offered the job.

- Academic Training (0-20 punts/ points) - Degree in physics, computational neuroscience or similar demonstrable training would be valued - Master in physics, computational neuroscience or similar would be valued.
- Other professional training and experience, adequacy to the proposed profile (0-60 punts/ points): Previous experience:
- Stochastic Control Theory for Motor Control • Recurrent Neural Networks • Analysis of Behavioural Data • Reinforcement Learning • Multi‑alternatives Decision Making
- Altres mèrits (0-20 punts) -Languages: High level of English -Programming: Matlab and Python. / -Idiomes: Nivell alt d'anglès -Programació: Matlab i Python.

Minimum punts: 70

📌 ENG-PSR-P-30: Specialist support technician: (Barcelona)
🏢 Pompeu Fabra University
📍 Barcelona