Byoung-Tak Zhang

Papers

Total Citations

H-Index

About

Byoung-Tak Zhang is a pioneering researcher at the intersection of artificial intelligence, human-robot interaction, and embodied cognition. His work focuses on enabling machines to understand and act upon natural language in physically grounded, real-world environments — a challenge that sits at the frontier of modern AI research. Zhang's most notable recent contributions include PROGrasp (2024), which advances Interactive Object Grasping by drawing on principles from linguistic pragmatics to enable more natural, contextually aware human-robot communication — moving beyond rigid category-based object specification toward genuinely conversational interaction. His Socratic Planner (2025) tackles the demanding problem of Embodied Instruction Following, introducing a self-questioning framework that enables zero-shot compositional task planning without reliance on supervised training or handcrafted examples. These works reflect Zhang's broader commitment to building AI systems that reason, communicate, and act more like humans — particularly in navigating complex, dynamic environments. While his most recent papers are still accumulating citations (4 and 3 respectively), their novelty and ambition signal growing influence in robotics and embodied AI communities. Students interested in grounded language understanding, pragmatic reasoning, or autonomous agents will find Zhang's research both intellectually rich and highly relevant to the field's most pressing open problems.