Tomasz Korbak
Papers
2
Total Citations
4
H-Index
2
About
Tomasz Korbak is a researcher at the intersection of cognitive science, artificial intelligence, and computational linguistics, whose work explores how symbolic communication emerges and how language models can be aligned with human values. His major contributions center on two key areas: the computational modeling of language evolution and the alignment of large language models (LLMs) with human preferences. In his highly cited work on "Models of symbol emergence in communication," Korbak provides a conceptual review and practical guide for avoiding local minima in simulations of communication emergence—a critical challenge for researchers studying the origins of language through multi-agent reinforcement learning and emergent communication. This work bridges cognitive science and machine learning, offering methodological insights for fields ranging from developmental psychology to robotics. Beyond this, Korbak has made notable contributions to reinforcement learning from human feedback (RLHF), developing techniques to better align LLMs with complex human values. His research has garnered significant attention, with his most cited papers accumulating hundreds of citations, reflecting the impact of his work on both theoretical foundations of communication and practical AI alignment.
Research Focus
Key Achievements
Top Papers
- 1
- 2