Boris Katz
Papers
8
Total Citations
49
H-Index
5
About
Boris Katz is a leading researcher at the intersection of natural language processing, robotics, and artificial intelligence, whose work focuses on enabling machines to understand and execute grounded language commands in the physical world. His major contributions center on developing compositional architectures—neural networks structured according to linguistic parse trees—that allow robots to systematically generalize from limited training data to novel, complex instructions. Katz’s 2021 paper on compositional networks for grounded language understanding (13 citations) demonstrates how deep networks can achieve human-like flexibility when interpreting new combinations of concepts, a critical step toward robust human-robot interaction. He has also pioneered the formalization of social reasoning for autonomous agents, extending Markov decision processes to incorporate rich social interactions from microsociology (7 citations). His work on temporal grounding graphs (2017) and natural-language-to-LTL semantic parsers (2020) has advanced robots’ ability to understand commands in context, accruing visual-linguistic information over time. With over 40 total citations across his most influential papers, Katz’s research bridges symbolic reasoning and deep learning, offering a principled path toward robots that can learn language as naturally as children do—through observation, context, and compositional understanding.
Research Focus
Key Achievements
Top Papers
- 1
- 2Partially Occluded Hands:7 citations · 2019
- 3Social Interactions as Recursive MDPs7 citations · 2021
- 4Incorporating Rich Social Interactions Into MDPs6 citations · 2022
- 5
- 6
- 7Compositional RL Agents That Follow Language Commands in Temporal Logic3 citations · 2021
- 8