Maxim Patratskiy

Moscow Institute of Physics and Technology

Papers

Total Citations

H-Index

About

Maxim Patratskiy is a researcher at the forefront of integrating large language models with robotics, specializing in task planning and visual feedback systems. His key research areas include embodied AI, instruction following, and replanning in dynamic environments. Patratskiy’s major contribution is the development of LERa (Look, Explain, Replan), a novel Visual Language Model-based approach that enables robots to adapt to real-world changes and failures by incorporating visual feedback into their planning processes. This work addresses a critical limitation in traditional LLM-based robotics, which rely solely on textual inputs. While his most-cited paper, "LERa: Replanning with Visual Feedback in Instruction Following" (2025), has garnered 2 citations in its early stage, it represents a significant step toward more robust and flexible autonomous systems. Patratskiy’s research promises to enhance the reliability of robots in unstructured environments, making them more responsive to unexpected obstacles and user corrections. His work is particularly notable for bridging the gap between language understanding and visual perception, offering a practical solution for real-world deployment of intelligent robotic agents.