I’m Zeynep, a psychologist holding a Master’s in Psychoanalytic Psychology from IPU Berlin. If psychoanalysis can be defined in a sentence, it is the study of how unconscious1 processes shape conscious experience. One of the main insights of psychoanalysis is that the conscious mind is not the “master in its own house”. Our actions as humans are often misaligned with our conscious goal statements. Initially, it was this question of how our unconscious, which I thought to be misaligned with the conscious mind2, impacts the LLMs we train that fueled my interest in alignment/safety research.
My background spans affective neuroscience, human factors research, and independent research. I’ve previously interned at Deep Blue — a European human factors company working with ESA and EASA — where I did work directly at the intersection of research and policy making relating to the implementation of AI assistants in safety-critical environments as well trust formation in human–AI teams in aviation contexts. I am a member of BLISS e.V., a machine learning organization in Berlin, and have previously been part of the outreach team of Knowing Neurons, an UCLA-based (neuro)science communication organization.
My master’s thesis dealt with the physiological separation of Panksepp’s SADNESS system into GRIEF and PANIC. My current research interests include creating a taxonomy of persona drift during long-context chatbot use, trust and shared interfaces in human-AI teams, LLM-based psychiatric diagnostics, and therapy aids.
In my free time, I can be found reading and writing sci-fi, learning German and Arabic, building side projects such as my dream repository, and studying the history of science.
Feel free to reach out to me on zeynep@zeynepaltayli.com, or on any of the social media I have linked below!