Hayden Shively
Papers
1
Total Citations
282
H-Index
1
About
Hayden Shively is a leading researcher in meta-reinforcement learning and robotics, best known for developing Meta-World, a benchmark that has fundamentally reshaped how multi-task and meta-learning algorithms are evaluated. His seminal 2019 paper, "Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learning," has garnered over 282 citations, reflecting its critical role in advancing the field. Shively identified a key limitation in existing meta-RL research—narrow task distributions that fail to test generalization—and responded by creating a diverse, open-source suite of robotic manipulation tasks. This benchmark enables researchers to rigorously assess how quickly and effectively algorithms can learn new skills by leveraging prior experience, moving beyond toy problems toward real-world applicability. His work has become a standard reference for evaluating multi-task and meta-learning approaches, influencing both academic research and practical robot learning. Through Meta-World, Shively has provided the community with a vital tool for accelerating progress toward more adaptable, sample-efficient AI systems.
Research Focus
Key Achievements
Top Papers
- 1