About me
I am a Ph.D. student in Computer Science at the Robotics and Perception Group (RPG), advised by Professor Davide Scaramuzza, and an Associated Researcher at the ETH AI Center. My research lies at the intersection of theoretical machine learning, sequence modeling, and the design of efficient neural network architectures. I am particularly interested in understanding the principles that govern learning, representation, and computation in sequence models, and in using these insights to develop architectures that are both theoretically grounded and practically effective. This work also connects to applications such as neuromorphic, event-based vision systems. More broadly, I am interested in dynamical systems, theoretical computer science, and applied mathematics, especially where these areas give rise to deep computational and mathematical challenges. If you would like to discuss my work or explore potential collaborations, please feel free to contact me or browse my recent papers in the Publications tab.
Before joining RPG, I obtained my Master's degree in Computer Science from the Faculty of Technical Sciences in Novi Sad, Serbia. During my Master's studies, I completed an internship at the University of Cambridge under the supervision of Professor Pietro Liò.
Outside of research, I enjoy competitive programming, mathematical puzzles and games, chess, and longevity-focused endurance sports, including running, brisk walking with Nordic poles, and rowing.
Recent News
-
Jun. 19, 2026
2 papers are accepted at ECCV 2026 in Malmö, Sweden: "Low-latency Event-based Object Detection with Spatially-Sparse Linear Attention" and "Hybrid Event Frame Sensors: Modeling, Calibration, and Simulation".
-
May. 25, 2026
I gave a talk on "Limits of Deep Learning" at the FLaNN Seminars - a series of weekly online seminars on Formal Language Theory, Natural Language Processing, Machine Learning and Computational Linguistics.
-
Apr. 10, 2026
"FastEventDGS: Deformable Gaussian Splatting for Fast Dynamic Scenes from a Single Event Camera" has been accepted at CVPR 2026 (Denver, Colorado).
-
Jan. 30, 2026
A big milestone for me! Our paper "A benchmark of expert-level academic questions to assess AI capabilities" has been published in Nature!
-
Jan. 26, 2026
"Maximizing Asynchronicity in Event-based Neural Networks" has been accepted at ICLR 2026 (this year in Brazil)!
-
Oct. 27, 2025
I was selected as a Top Reviewer for NeurIPS 2025. This recognition includes complimentary conference registration, awarded in acknowledgment of my contributions to the NeurIPS review process.
-
Sep. 29, 2025
The big project I worked on, titled "Humanity's Last Exam", appeared in the New York Times magazine.
-
Aug. 4, 2025
I gave a talk at Google Research and Google DeepMind New York titled "Breaking and Building: On the Computational Limits and Provable Stability of Modern Sequence Models".
-
Apr. 16, 2025
"Perturbed State Space Feature Encoders for Optical Flow with Event Cameras" has been accepted at CVPRW 2025 in Nashville, Tennessee. Paper is available.
-
-
Feb. 6, 2025
My work on Limitations of Deep Learning appeared in QuantaMagazine! Article is available here.
-
Jan. 22, 2025
"Limits of Deep Learning: Sequence Modeling through the Lens of Complexity Theory" has been accepted at ICLR 2025 (this year in Singapore)!
-
Aug. 1, 2024
I gave a 1.5-hour talk on "Limits of Deep Learning: Sequence Modeling through the Lens of Complexity Theory" at Cohere for AI. Details here.
-
Apr. 11, 2024
I gave a 1-hour talk on "State Space Models for Event Cameras - New perspectives and directions" at KTH Royal Institute of Technology, organised by Prof. Jörg Conradt. Details here.
-
Mar. 8, 2024
I gave a 1-hour talk on "State Space Models for Event Cameras" at Stochastic Finance Group led by Prof. Josef Teichmann. Details here.
-
Feb. 26, 2024
"State Space Models for Event Cameras" has been accepted at CVPR 2024 in Seattle, as a Spotlight Paper!
