Ramin Hasani
Papers
15
Total Citations
393
H-Index
10
About
Ramin Hasani is a pioneering researcher at the intersection of biologically inspired neural networks, safe robot learning, and autonomous systems. Drawing inspiration from the elegant neural architecture of the nematode *C. elegans*, Hasani has championed the development of liquid neural networks — a class of adaptive, interpretable recurrent models whose time-varying dynamics yield remarkable robustness and generalization. His foundational 2019 work on worm-inspired neural networks laid the groundwork for subsequent breakthroughs, including liquid neural networks that enable robust autonomous flight navigation in out-of-distribution environments (79 citations). A central theme of his research is safety-critical AI: his BarrierNet framework (99 citations) introduced differentiable control barrier functions that guarantee the safety of learned robot controllers end-to-end. Hasani has also advanced human-robot collaboration through biosignal-based supervisory control systems and explored the boundaries of model-based reinforcement learning for autonomous racing. Across his portfolio, he consistently bridges theoretical rigor with real-world deployment, producing work that is both deeply principled and immediately applicable — making him a compelling figure for students pursuing safe, intelligent, and interpretable autonomous systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2Robust flight navigation out of distribution with liquid neural networks79 citations · 2023
- 3Designing Worm-inspired Neural Networks for Interpretable Robotic Control43 citations · 2019
- 4Latent Imagination Facilitates Zero-Shot Transfer in Autonomous Racing35 citations · 2022
- 5
- 6
- 7
- 8
- 9Adversarial Training is Not Ready for Robot Learning12 citations · 2021
- 10