PD Dr. Markus J. Hofmann

I always had a strong fascination for all types computational models of cognition. I did connectionist and language models long before the triumphal procession of artificial intelligence. Recently, I became enthused about predictive modeling and all types of machine learning models. Therefore, I decided to found the Center for Neurocognitive Modeling for the institute of psychology of the BUW as a special interest group bringing together all researchers, who like model-based cognitive neuroscience. All collaborators and PhD student are cordially invited to join the CNM.

I did my PhD with Arthur Jacobs in 2011, where my fascination for symbolic representations in cognitive process models emerged. We included language models into our localist connectionist model, the AROM, to account for association ratings, recognition memory and primed lexical decisions. In 2021, I finished my habilitation on the neurocognitive dynamics of semantic memory processes in algorithmic models. Not only due to my studies of Philosophy with a focus on epistemology, countless collaborations with physicians, physicists and neurolinguists, but also due to Philosopher as a PhD student I am widely open to interdisciplinary work. I also supervised a PhD thesis of a computational biologist, which helped me to professionalize my own programming skills. I have a long-standing collaboration with Chris Biemann, director of the hub of computing and data science of the University of Hamburg, with whom I investigated the performance of all types of language models in predicting human performance.

In addition to computational methods on the theoretical level, I really like advanced statistical and particularly neuroscientic analysis models. I started my career as an EEG researcher and then went further by learning and teaching fMRI methods. To measure the hemodynamic response in an ecologically more valid setting, I use near-infrared spectroscopy, which allows to track cortical blood consumption by light absorption. We also combined this method with eye tracking in order to examine hemodynamic responses that are elicited by extremely rapidly paced events, such as eye fixations during reading. But I also adore the appeal of easily interpretable behavioral data. 

My latest major research interest are individual text corpora, which we combine with language and predictive models in order to predict personality, knowledge and intelligence.

• Curriculum Vitae (September 2023)
• Grants (September 2024)