Bariscan (Baris) Kurtkaya

I am a prospective PhD student in Electrical and Computer Engineering at the University of California, Santa Barbara, where I will be advised by Prof. Nina Miolane as a member of the Geometric Intelligence Lab.

I am currently a KUIS AI Fellow and a second-year MSc student in Computer Science and Engineering at Koç University, where I am fortunate to be advised by Prof. Yücel Yemez. Alongside my graduate studies, I have also worked as a visiting scholar at Stanford University, contributing to research in computational neuroscience and memory dynamics within the Schnitzer Group under the advisement of Prof. Mark Schnitzer and Dr. Fatih Dinç.

My research focuses on understanding the geometry and dynamics of memory in both biological and artificial systems. I am particularly interested in the intersection of machine learning, neuroscience, and dynamical systems theory.

Throughout my academic journey, I've had the opportunity to collaborate with researchers at several institutions, including Stanford University, Harvard University, UCSB, Virginia Tech, Washington University in St. Louis, and the University of Milan. I feel incredibly fortunate to have met and worked with such brilliant, kind, and passionate people from around the world.

Beyond academics, I am deeply passionate about long-distance running, cinematography, photography, poetry, and music of all kinds. I'm currently learning to play the piano and always looking for new ways to express myself creatively.

Email  /  Google Scholar  /  CV  /  GitHub  /  LinkedIn

Selected Publications

Dynamical phases of short-term memory mechanisms in RNNs

We investigate the scaling relationship between the learning rate, length of delay interval and the dynamical phases underlying short-term memory mechanisms in recurrent neural networks (RNNs), focusing on structures such as limit cycles and slow-point manifolds.

RAVE: Randomized Noise Shuffling for Fast and Consistent Video Editing with Diffusion Models

We propose a latent grid and shuffling framework that enables text-based zero-shot video editing by leveraging pretrained text-to-image diffusion models.

Design and source code from Jon Barron's website