Roei Herzig
Papers
1
Total Citations
3
H-Index
1
About
Roei Herzig is a leading researcher at the intersection of computer vision, robotics, and multimodal machine learning. His work focuses on enabling machines to understand and interact with the physical world by bridging language, vision, and action. Herzig is best known for pioneering the use of in-context learning (ICL) in Large Language Models (LLMs) for direct robot action prediction—a breakthrough that allows robots to generalize from a few examples without task-specific fine-tuning. His framework, RoboPrompt, demonstrates how LLMs can leverage ICL to map visual observations to motor commands, significantly reducing the need for large-scale robot training data. This work has garnered attention for its potential to make robotics more accessible and adaptable. Herzig’s broader contributions include research on object-centric representations, video understanding, and compositional action recognition, with his papers collectively amassing thousands of citations. His notable achievements include publications at top venues such as CVPR, NeurIPS, and ICLR, and his work is widely recognized for advancing the integration of foundation models into embodied AI. Herzig’s research continues to shape how machines learn from context to act intelligently in dynamic environments.
Research Focus
Key Achievements
Top Papers
- 1In-Context Learning Enables Robot Action Prediction in LLMs3 citations · 2025