Brain with neural Laptop

Marcel Binz Receives ERC Starting Grant for MIDAS

AI Awards & Grants Computational Health HCA

Dr. Marcel Binz from the Institute for Human-Centered AI at Helmholtz Munich has received a Starting Grant from the European Research Council (ERC). Through the MIDAS project, he will develop large-scale computational models designed to simulate, predict and explain human behavior across a wide range of tasks and situations. The project will receive approximately €1.5 million in funding over five years.

In its latest funding round, the European Research Council is awarding a total of €705 million to 421 researchers across Europe. ERC Starting Grants support researchers at the beginning of their scientific independence, enabling them to pursue ambitious projects and establish their own research teams. Each grant provides up to €1.5 million over a maximum period of five years. A total of 4,807 proposals were submitted in the latest funding round, of which 8.8% were selected for funding.

MIDAS: Modelling Human Behaviour at Scale

How does the human brain process information and translate it into behavior? Marcel Binz, Postdoctoral Researcher at the Institute for Human-Centered AI at Helmholtz Munich, addresses this fundamental question in cognitive science through his project “Modeling the Integrated minD At Scale” (MIDAS). On 1 October 2026, Binz will join the Max Planck Institute for Biological Cybernetics in Tübingen as a Group Leader.

From Individual Tasks to Integrated Models

Computational cognitive models translate psychological theories into algorithms. In doing so, they can help researchers investigate cognitive processes that cannot be directly inferred from observable behaviour alone. To date, however, such models have generally been tailored to individual tasks or specific research domains. This makes it difficult to integrate findings from different experiments and generalize them to new situations.

MIDAS therefore aims to develop large-scale, integrated models of human cognition. These models will be designed to simulate and predict human behaviour across a range of experimental tasks while also providing interpretable explanations of the underlying cognitive processes. To achieve this, Binz combines approaches from cognitive science, machine learning and language modelling.

A key foundation of the project is Psych-101, a dataset that brings together more than ten million human decisions from 160 experiments in a standardized format. Building on this resource, Binz aims to investigate which cognitive processes operate across different tasks, which are specific to particular situations, and how they interact to generate human behavior.

Cognitive Models for More Naturalistic Settings

MIDAS will also gradually extend cognitive modelling beyond controlled experiments to more naturalistic settings. The planned research includes studies of human conversations as well as complex visual and computer-based tasks. In the long term, the findings could provide a foundation for future applications in fields such as education and psychiatry.

About the European Research Council

The European Research Council was established by the European Union in 2007 and funds excellent frontier research across Europe. It supports creative researchers of all nationalities and at different career stages. Its core funding schemes include Starting Grants, Consolidator Grants, Advanced Grants and Synergy Grants. The ERC forms part of Horizon Europe, the EU’s framework programme for research and innovation.

Dr. Marcel Binz

Postdoc

Related news

Digital brain interface activated by a man, highlighting the connection between human cognitive abilities and artificial intelligence in a hightech, futuristic environment

AI, New Research Findings, Computational Health, HCA,

AI That Thinks Like Us – and Could Help Explain How We Think

Researchers at Helmholtz Munich have developed an artificial intelligence model that can simulate human behavior with remarkable accuracy. The language model, called Centaur, was trained on more than ten million decisions from psychological…