Sujin Jeon
Papers
1
Total Citations
3
H-Index
1
About
Sujin Jeon is a rising researcher in embodied AI and natural language processing, with a focus on enabling robots to follow complex human instructions in interactive environments. Her work addresses the critical challenge of compositional task planning for Embodied Instruction Following (EIF), where agents must navigate and manipulate objects based on natural language commands. Jeon's most notable contribution is the "Socratic Planner," a zero-shot planning framework that leverages self-question answering to decompose tasks without requiring supervised training data or few-shot examples. This innovative approach, presented in 2025, has already garnered attention with 3 citations in its early release, signaling its potential impact on the field. By eliminating the need for expensive annotated datasets, Jeon's method makes embodied AI more scalable and accessible. Her research sits at the intersection of robotics, planning, and language understanding, pushing toward more autonomous and adaptable agents. As a young scholar, Jeon is establishing herself as a key voice in zero-shot learning for embodied systems, with her work promising to influence how future robots interpret and execute open-ended instructions in dynamic, real-world settings.
Research Focus
Key Achievements
Top Papers
- 1