Deep learning (DL) has recently achieved remarkable success due to the continuous growth in model sizes. However, this growth has led to increased energy consumption. Hardware implementation of digital DL can help reduce energy usage, but the Von Neumann architecture of current DL has hindered its practical realization. In contrast, the brain exhibits energy-efficient multiscale spatiotemporal processing. Biologically plausible (BiP) frameworks have emerged as alternatives to mainstream DL. These methods use bottom-up and top-down signals, incorporating feedforward and feedback mechanisms, and local objectives instead of global error. Recently, I demonstrated that a BiP opto-analog hardware can achieve competitive performance compared to digital DL for feedforward networks. However, transformers, the backbone of current DL, are challenging to implement due to the input-dependent quadratic complexity in the transformer's attention. This project leverages the multiscale dynamics in the primary vision system to explore BiP architectures for transformers. The project is hosted at the University of Tübingen under Matthias Bethge and Thomas Euler, who have a long-standing effort in the system identification of mouse retina via DL. The project has three objectives. First, I will extract top-down information from neural recordings of ganglion cells in the mouse retina, focusing on unique spatiotemporal features that maximally activate specific cell types. Next, I will combine top-down signals with bottom-up models of the retina using recurrent architectures with linear complexity and compare their performance in classification tasks to a vision transformer for the retina. Lastly, I propose a BiP transformer with local weight updates. I will examine the robustness of models under data distribution shifts and noise injection. A positive outcome of the project will address energy and cost issues of AI and help me progress my academic career in this interdisciplinary field.

The most important open questions in European paleoanthropology concern the timing, number and origin of early human dispersals into the continent, the identity and number of taxa present, and their possible interactions. These issues remain unresolved partly due to the lack of research in South-East Europe, a region at the crossroads between continents and a refugium for fauna, flora and possibly human populations during glacial times. The PI’s previous work there (ERC StG PaGE) aimed to add new evidence to further our understanding of human evolution on the continent. PaGE led to the discovery of several new Paleolithic sites, including the oldest radiometrically dated archaeological site from South-East Europe, placing the region squarely in the Paleolithic map of Europe. CROSSROADS is an ambitious, groundbreaking research program that builds on the foundation of PaGE to take Paleolithic research in the region to a new level. In contrast to the exploratory goals of PaGE, it focuses on the early part of the Paleolithic targeting the following questions: Can human presence in South-East Europe, considered a major dispersal route into the continent from Africa and the Near East, be documented beyond the current known chronology of ca. 500 ka BP, as shown in the West? Is there a gap between the earliest human arrival and subsequent human activity in the region, or was human habitation continuous, as would be expected in a refugium? What was the environmental backdrop of the early human dispersal and subsequent evolution? How did behavioral / biological change correlate with environmental changes, chronology and landscape use? Can we document a higher level of diversity in the human fossil record of the region than would be expected under evolutionary scenarios developed on Western European evidence, suggesting that different evolutionary processes were at work? The answers to these questions will be essential for testing hypotheses about human evolution in Eurasia.

In antiquity and the early Middle Ages, a network of trade routes known as the Silk Road connected east Asia and the Мiddle East. The Silk Road was not just an economic link, but also the avenue for cultural and even genetic exchanges between these regions. Recent genetic discoveries have hinted that such connections might have begun much earlier, during the Pleistocene. The Pleistocene period is of fundamental importance for human history. It is then that our ancestors evolved and colonised the entire Old World, surviving a suite of major extinction events – and they did so against a dramatic backdrop of ice ages and warmer interglacial phases which substantially altered their habitats. Conquering the extreme environments of arid central Asia to eventually settle the entire Asian mainland and beyond is one of the most impressive feats in this story. Unfortunately, there are too few known Pleistocene archaeological sites in central Asia to allow us to piece together when and how this happened. PALAEOSILKROAD will resolve this deficit by surveying central Asian mountain foothills as both corridors for human and animal movements and archives of past climate change. The project will discover new sites in the Tian Shan, Dzungar, and southern Altai foothills (Kazakhstan) and use them to examine if and how 1) humans were able to survive in the foothills throughout the last glacial cycle (110-11 500 years ago), and 2) periodic advances of mountain glaciers motivated dispersals, population segmentation, and behavioural adaptations. To address these questions, PALAEOSILKROAD will take an ambitious approach rooted in archaeology and contextualised by palaeoenvironmental reconstruction. The results of this project will change the way we understand human dispersals on a global scale and the resilience of early humans in the face of environmental challenges, providing a major missing link to explain how Homo sapiens became the only surviving species of our genus.

While in many countries women have been increasing their participation in higher education, fields of study continue to be segregated by gender. Men dominate the technical fields which include mathematics, engineering and the physical sciences, whereas women tend to engage in areas that have a social and caring dimension. Despite numerous policies to enhance gender equalities in society, gender segregation in higher education, ironically, appears to be more pronounced in post-industrial countries, such as Germany, than in developing/transforming societies. Although gender differences in field of study and occupational choices are shaped by multifarious influences, the role of parents might be particularly worth attention because parents are the primary and one of the most influential socialising agents in childhood and adolescence. In fact, one possible source of students’ knowledge about and interest in a particular field of study and career early on stems from parental education and occupation. The impact of parents’ occupational fields and its gender typicality has received less attention than parental education in early research. Even though some studies have identified the association of parental occupations with students’ engagement in gender-atypical fields of study, they do not provide any empirical evidence of the mechanism for the relationship. Despite the possibly significant influence of parental occupation on the educational and career choices of adolescent boys and girls, previous quantitative studies rarely examined the mechanisms for the relationship between such gender differences and parental occupations. With the proposed research “Understanding the impact of parental occupation on gender differences in field of study and occupational choices in Germany”, that comprises 3 projects using secondary and primary data, I seek to offer new evidence on how parental occupations influence students’ choices of a gender-atypical field of study and career.