Papers

15

Total Citations

2,287

H-Index

10

About

Shixiang Gu is a prominent researcher at the intersection of deep reinforcement learning (RL) and robotics, whose work has fundamentally advanced how autonomous systems acquire complex behavioral skills. His most influential contribution, "Deep Reinforcement Learning for Robotic Manipulation with Asynchronous Off-Policy Updates" (2017), garnered over 1,450 citations and demonstrated how asynchronous, sample-efficient RL methods could enable robots to learn manipulation tasks with minimal human intervention. Gu has consistently tackled one of the field's central challenges — sample inefficiency — through innovations like model-based acceleration in continuous deep Q-learning (337 citations) and deployment-efficient offline optimization. His work on hierarchical reinforcement learning (265 citations) extended RL's reach to more complex, real-world tasks, while his "Leave No Trace" framework introduced principled approaches for safe, autonomous environment resetting during training. More recently, Gu has explored unsupervised off-policy learning to develop reward-free robotic skills and applied data-driven methods to household robot challenges. Across his career, his research has collectively shaped modern robotic learning pipelines, emphasizing autonomy, safety, and practical deployability — making him a highly influential figure for students and practitioners in AI-driven robotics.

Research Focus

Key Achievements

10
H-Index
15
Papers
2,287
Total Citations
152
Avg Citations/Paper
🏆 Most Cited Paper
Deep reinforcement learning for robotic manipulation with asynchronous off-policy updates
1,452 citations · 2017
📈 Most Prolific Year: 2020 (3 Papers)
🤝 Key Collaborators: 42
🏛 Institutions: University of Cambridge, Google (United States), Max Planck Society, The University of Tokyo, Shibuya (Japan)

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 13 days ago