Tabish Rashid

Papers

Total Citations

H-Index

About

Tabish Rashid is a rising researcher at the forefront of artificial intelligence, specializing in imitation learning, generative models, and the scaling of embodied agents. His work bridges the gap between generative AI and sequential decision-making, with a focus on how models can replicate complex human behavior. In his highly cited 2023 paper, "Imitating Human Behaviour with Diffusion Models," Rashid pioneered the use of diffusion models—traditionally dominant in text-to-image generation—as observation-to-action models for sequential environments. This work addresses the stochastic, multimodal nature of human behavior, offering a powerful new framework for imitation learning. More recently, his 2024 study on "Scaling Laws for Pre-training Agents and World Models" explores how increasing model parameters, dataset size, and compute improves the performance of embodied agents across domains like robotics and video games. By demonstrating that generative pre-training objectives can effectively model agent behavior, Rashid is helping to establish foundational principles for scaling intelligent systems. With his papers already garnering significant attention in the AI community, Tabish Rashid is shaping the next generation of agents that learn from and interact with the world.