Atalay Donat

Papers

Total Citations

H-Index

About

Atalay Donat is a rising researcher in artificial intelligence, with a primary focus on advancing imitation learning and state-space models for robotics and sequential decision-making. His most notable contribution is the introduction of MaIL (Mamba Imitation Learning), a novel architecture that leverages the Mamba state-space model as a powerful alternative to Transformer-based policies. By designing Mamba to selectively focus on key data features, Donat addresses critical limitations in computational efficiency and long-range dependency modeling, offering a more scalable solution for imitation learning tasks. Although his seminal 2024 paper has already garnered 3 citations in its early stages, signaling growing interest from the AI community, Donat’s work represents a significant step toward more efficient and robust policy learning. His research sits at the intersection of imitation learning, sequence modeling, and robotics, promising to reduce the computational overhead of current state-of-the-art methods while maintaining high performance. As an emerging scholar, Donat is poised to make lasting contributions to the field, particularly in developing lightweight, high-fidelity models for real-world robotic applications.