ERC Starting Grants are annually awarded to talented young scientists, with over two and up to seven years of experience since completion of PhD. The recipients at Radboud University will be doing research into energy efficient data storage, history of impossible numbers, mechanisms of memory in the brain, verification methods for artificial intelligence, motivations for bullying and a GPS in the brain.
Antiferromagnetic Spin Transport with Relativistic Waves
Dmytro Afanasiev, Institute for Molecules and Materials (IMM)
Modern data centres are running into the limits of their performance because of their growing energy demand. Therefore, there is a need to develop new ways and technologies for faster and more energy-efficient data processing and storage. Photonics provide more energy-efficient and faster transfer of data. However, processing data at the same rates still is a major challenge. Magnonics is widely seen as one of the most appealing solutions to this problem. Instead of light waves, magnonics exploits spin waves– the collective propagating precession of spins in magnetic materials. However, there are still fundamental key mechanisms that need to be tackled. For example, how can magnonics be pushed into the THz domain and enter the nonlinear regime?
Afanasiev and his research team aim to tackle this by generating ultrashort large amplitude spinwaves pulses that can zip undisturbed over long distances unlocking the nonlinear regime of interaction between the pulses, other spinwaves and even macroscopic spin textures. Although the idea is fundamental in nature, it will create new paths to ground-breaking new computing technologies. Read more
The impossible and the imaginable: late-medieval semantics of impossibilities and the roots of complex mathematics
Graziana Ciola, Radboud Center for the History of Philosophy and Science
Try to imagine a round square. This is impossible. In medieval logic, the “Chimera” is equivalent to the round square: an inconceivable absolute impossibility. For a long time in history, philosophers and scientists held that we only conceive what is in some sense possible. Therefore, premodern mathematics dismissed the square root of a negative number as the impossible result of an impossible operation. Yet by the 16th century these "impossible numbers" had become conceivable and manipulable. How and why did this shift happen?
Graziana Ciola argues the medieval concept Chimera is the grandmother of imaginary numbers. Her research project shows how medieval logic widened the limits of what we can conceive, with revolutionary consequences on how we understand reality. Ciola's research project uses a key-concept analysis and the method of historical and rational reconstruction to a rich textual corpus.
Cracking the Synaptic Memory Code
Anne-Sophie Hafner, Donders Institute for Brain, Cognition and Behaviour
Synapses play an important role in long-term information storage in the brain. They are highly dynamic: in the adult mouse brain, it takes a few days for dendritic spines to be replaced. Similarly at the molecular level, most synaptic proteins have half-lives in the order of a week, meaning they constantly need to be replaced by freshly produced ones. Understanding how long-term memory can arise from unstable elements is one of today’s great neuroscience challenges.
Anne-Sophie Hafner discovered that most synapses produce their own proteins locally. She will combine multiple research methods to unravel how local production of new proteins contributes to information storage at synapses. Such a fundamental understanding of brain function is needed to provide new avenues of defense against neurodegenerative diseases.
Data-Driven Verification and Learning Under Uncertainty
Nils Jansen, Institute for Computing and Software Science (ICIS)
Artificial intelligence (AI) is entering our everyday lives, with applications in fields like healthcare, transportation, finance, or robotics. Many of those fields require strong safety requirements on the AI systems that are used. A particular AI, or machine learning, technique is a reinforcement learning, which generally learns to behave optimally via trial and error. Consequently, and despite its huge success in the past years, reinforcement learning generally lacks mechanisms to constantly ensure safe behavior. On the other hand, formal verification is a research area that aims at providing formal guarantees on a system’s correctness and safety, based on rigorous methods and precise specifications. However, fundamental challenges obstruct the effective application of verification to reinforcement learning so far.
The main objective of Nils Jansen’s research project is to develop novel and data-driven verification methods that tightly integrate with reinforcement learning. In particular, he will develop techniques that address real-world challenges to the safety of AI systems in general: Scalability, expressiveness, and robustness against the uncertainty that occurs when operating in the real world. The overall goal is to advance the real-world deployment of reinforcement learning.
Motivations at the Automatic and Deliberate level to bully
Tessa Lansu, Behavioral Science Institute (BSI)
Bullying among youth is a major problem with devastating consequences for victims. But why do youth engage in or support bullying? Research has mainly focused on a deliberate motivation to strive for higher status. Tessa Lansu proposes that not only the motivation to ‘be on top’, but also the motivation to ‘not dangle at the bottom’ plays an important role in bullying.
Lansu aims to also examine motivation at the automatic level. She will study both the reward value of high status and the threat value of low status, on the deliberate level as well as on the automatic level. The insights from this project will change our understanding of why youth engage in or support bullying. Moreover, the study will aid the development of anti-bullying and anti-following programs and interventions.
The Body Positioning System: A GPS for somatosensory space
Luke Miller, Donders Institute for Brain, Cognition and Behaviour
If an object touches the body, we reach for it. This is so commonplace that we hardly give it a second thought. However, this “simple” behavior is actually an incredibly complicated problem the brain must solve on a daily basis. How can the brain accurately pinpoint where the object is in three-dimensional space and move the reaching hand towards it?
Luke Miller introduces the first neurocomputational framework aimed at solving this mystery. He proposes that the brain uses a somatosensory version of Global Positioning Systems (GPS). GPS pinpoints an object on Earth by calculating its distance from multiple satellites. Likewise, the brain reduces localization to its geometry. Miller’s innovative proposal provides the first neurocomputational model of tactile localization. This could help us in making progress, for instance in designing prosthetics that can be embodied.
In total, 408 researchers have won this year's ERC Starting Grants. Read more on the website of ERC.