Workshop
Safe and Robust Control of Uncertain Systems
Ashwin Balakrishna 路 Brijen Thananjeyan 路 Daniel Brown 路 Marek Petrik 路 Melanie Zeilinger 路 Sylvia Herbert
Mon 13 Dec, 8 a.m. PST
Control and decision systems are becoming a ubiquitous part of our daily lives, ranging from serving advertisements or recommendations on the internet to controlling autonomous physical systems such as industrial equipment or robots. While these systems have shown the potential for significantly improving quality of life and industrial efficiency, the impact of the decisions made by these systems can also cause significant damages. For example, an online retailer recommending dangerous products to children, a social media platform serving content which polarizes society, or a household robot/autonomous car which collides with surrounding humans can all cause significant direct harm to society. These undesirable behaviors not only can be dangerous, but also lead to significant inefficiencies when deploying learning-based agents in the real world. This motivates developing algorithms for learning-based control which can reason about uncertainty and constraints in the environment to explicitly avoid undesirable behaviors. We believe hosting a discussion on safety in learning-based control at NeurIPS 2021 would have far-reaching societal impacts by connecting researchers from a variety of disciplines including machine learning, control theory, AI safety, operations research, robotics, and formal methods.
Schedule
Mon 8:00 a.m. - 8:15 a.m.
|
Introduction
(
Short Talk
)
>
SlidesLive Video |
Daniel Brown 馃敆 |
Mon 8:15 a.m. - 8:40 a.m.
|
Ye Pu
(
Invited Talk
)
>
SlidesLive Video |
Ye Pu 馃敆 |
Mon 8:40 a.m. - 8:45 a.m.
|
Ye Pu
(
Talk Q/A Session
)
>
|
Ye Pu 馃敆 |
Mon 8:45 a.m. - 9:10 a.m.
|
Aleksandra Faust
(
Invited Talk
)
>
SlidesLive Video |
Aleksandra Faust 馃敆 |
Mon 9:10 a.m. - 9:15 a.m.
|
Aleksandra Faust
(
Talk Q/A Session
)
>
|
Aleksandra Faust 馃敆 |
Mon 9:15 a.m. - 9:40 a.m.
|
Shie Mannor
(
Invited Talk
)
>
SlidesLive Video |
Shie Mannor 馃敆 |
Mon 9:40 a.m. - 9:45 a.m.
|
Shie Mannor
(
Talk Q/A Session
)
>
|
Shie Mannor 馃敆 |
Mon 9:45 a.m. - 9:50 a.m.
|
Learning Contraction Policies from Offline Data
(
Oral
)
>
SlidesLive Video |
Navid Rezazadeh 路 Negar Mehr 馃敆 |
Mon 9:50 a.m. - 9:55 a.m.
|
Safety-guaranteed trajectory planning and control based on GP estimation for unmanned surface vessels
(
Oral
)
>
SlidesLive Video |
Shuhao Zhang 路 Seth Siriya 馃敆 |
Mon 9:55 a.m. - 10:00 a.m.
|
Efficiently Improving the Robustness of RL Agents against Strongest Adversaries
(
Oral
)
>
SlidesLive Video |
Yongyuan Liang 路 Yanchao Sun 路 Ruijie Zheng 路 Furong Huang 馃敆 |
Mon 10:00 a.m. - 11:00 a.m.
|
Safe RL Panel Discussion
(
Discussion Panel
)
>
|
Animesh Garg 路 Marek Petrik 路 Shie Mannor 路 Claire Tomlin 路 Ugo Rosolia 路 Dylan Hadfield-Menell 馃敆 |
Mon 11:00 a.m. - 12:00 p.m.
|
Poster Session I ( Poster Session I ) > link | 馃敆 |
Mon 12:00 p.m. - 12:25 p.m.
|
Rohin Shah
(
Invited Talk
)
>
SlidesLive Video |
Rohin Shah 馃敆 |
Mon 12:25 p.m. - 12:30 p.m.
|
Rohin Shah
(
Talk Q/A Session
)
>
|
Rohin Shah 馃敆 |
Mon 12:30 p.m. - 12:55 p.m.
|
Angelique Taylor
(
Invited Talk
)
>
SlidesLive Video |
Angelique Taylor 馃敆 |
Mon 12:55 p.m. - 1:00 p.m.
|
Angelique Taylor
(
Talk Q/A Session
)
>
|
Angelique Taylor 馃敆 |
Mon 1:00 p.m. - 1:25 p.m.
|
Ugo Rosolia
(
Invited Talk
)
>
SlidesLive Video |
Ugo Rosolia 馃敆 |
Mon 1:25 p.m. - 1:30 p.m.
|
Ugo Rosolia
(
Talk Q/A Session
)
>
|
Ugo Rosolia 馃敆 |
Mon 1:30 p.m. - 2:30 p.m.
|
Safe RL Debate
(
Debate
)
>
|
Sylvia Herbert 路 Animesh Garg 路 Emma Brunskill 路 Aleksandra Faust 路 Dylan Hadfield-Menell 馃敆 |
Mon 2:30 p.m. - 2:35 p.m.
|
Reinforcement Learning with Feedback from Multiple Humans with Diverse Skills
(
Oral
)
>
SlidesLive Video |
Taku Yamagata 路 Ryan McConville 路 Raul Santos-Rodriguez 馃敆 |
Mon 2:35 p.m. - 2:40 p.m.
|
What Would the Expert do()?: Causal Imitation Learning
(
Oral
)
>
SlidesLive Video |
Gokul Swamy 路 Sanjiban Choudhury 路 James Bagnell 路 Steven Wu 馃敆 |
Mon 2:40 p.m. - 2:45 p.m.
|
Who Is the Strongest Enemy? Towards Optimal and Efficient Evasion Attacks in Deep RL
(
Oral
)
>
SlidesLive Video |
Yanchao Sun 路 Ruijie Zheng 路 Yongyuan Liang 路 Furong Huang 馃敆 |
Mon 2:45 p.m. - 3:45 p.m.
|
Poster Session II ( Poster Session II ) > link | 馃敆 |
Mon 3:45 p.m. - 4:00 p.m.
|
Closing Remarks
(
Short Talk
)
>
SlidesLive Video |
Brijen Thananjeyan 馃敆 |
-
|
Learning Behavioral Soft Constraints from Demonstrations
(
Poster
)
>
|
Arie Glazier 路 Andrea Loreggia 路 Nicholas Mattei 路 Taher Rahgooy 路 Francesca Rossi 路 Brent Venable 馃敆 |
-
|
Conservative and Adaptive Penalty for Model-Based Safe Reinforcement Learning
(
Poster
)
>
|
Jason Yecheng Ma 路 Andrew Shen 路 Osbert Bastani 路 Dinesh Jayaraman 馃敆 |
-
|
Execute Order 66: Targeted Data Poisoning for Reinforcement Learning via Minuscule Perturbations
(
Poster
)
>
|
Harrison Foley 路 Liam Fowl 路 Tom Goldstein 路 Gavin Taylor 馃敆 |
-
|
Learning Robustly Safe Output Feedback Controllers from Noisy Data with Performance Guarantees
(
Poster
)
>
|
Luca Furieri 馃敆 |
-
|
State Augmented Constrained Reinforcement Learning: Overcoming the Limitations of Learning with Rewards
(
Poster
)
>
|
Miguel Calvo-Fullana 路 Santiago Paternain 路 Alejandro Ribeiro 馃敆 |
-
|
Safe Learning of Linear Time-Invariant Systems
(
Poster
)
>
|
Farhad Farokhi 路 Alex Leong 路 Mohammad Zamani 路 Iman Shames 馃敆 |
-
|
MESA: Offline Meta-RL for Safe Adaptation and Fault Tolerance
(
Poster
)
>
|
Michael Luo 路 Ashwin Balakrishna 路 Brijen Thananjeyan 路 Suraj Nair 路 Julian Ibarz 路 Jie Tan 路 Chelsea Finn 路 Ion Stoica 路 Ken Goldberg 馃敆 |
-
|
Risk Sensitive Model-Based Reinforcement Learning using Uncertainty Guided Planning
(
Poster
)
>
|
Stefan Radic Webster 馃敆 |
-
|
Adversarial Training Blocks Generalization in Neural Policies
(
Poster
)
>
|
Ezgi Korkmaz 馃敆 |
-
|
Robust Physical Parameter Identification through Global Linearisation of System Dynamics
(
Poster
)
>
|
Yordan Hristov 路 Subramanian Ramamoorthy 馃敆 |
-
|
Specification-Guided Learning of Nash Equilibria with High Social Welfare
(
Poster
)
>
|
Kishor Jothimurugan 路 Suguman Bansal 路 Osbert Bastani 路 Rajeev Alur 馃敆 |
-
|
Safe Reinforcement Learning for Grid Voltage Control
(
Poster
)
>
|
Thanh Long Vu 路 Sayak Mukherjee 路 Renke Huang 路 Qiuhua Huang 馃敆 |
-
|
Distributionally robust chance constrained programs using maximum mean discrepancy
(
Poster
)
>
|
Yassine Nemmour 路 Bernhard Sch枚lkopf 路 Jia-Jie Zhu 馃敆 |
-
|
Unbiased Efficient Feature Counts for Inverse RL
(
Poster
)
>
|
Gerard Donahue 路 Brendan Crowe 路 Marek Petrik 路 Daniel Brown 馃敆 |
-
|
Parametric-Control Barrier Function-based Adaptive Safe Merging Control for Heterogeneous Vehicles
(
Poster
)
>
|
Yiwei Lyu 路 John Dolan 馃敆 |
-
|
Bayesian Inverse Constrained Reinforcement Learning
(
Poster
)
>
|
Dimitris Papadimitriou 路 Daniel Brown 路 Usman Anwar 馃敆 |
-
|
ProBF: Probabilistic Safety Certificates with Barrier Functions
(
Poster
)
>
|
Sulin Liu 路 Athindran Ramesh Kumar 路 Jaime Fisac 路 Ryan Adams 路 Peter J Ramadge 馃敆 |
-
|
Behavior Policy Search for Risk Estimators in Reinforcement Learning
(
Poster
)
>
|
Elita Lobo 路 Marek Petrik 路 Dharmashankar Subramanian 馃敆 |
-
|
Safe Online Exploration with Nonlinear Constraints
(
Poster
)
>
|
Eleanor Quint 路 Garrett Wirka 路 Stephen Scott 馃敆 |
-
|
Avoiding Negative Side Effects by Considering Others
(
Poster
)
>
|
Parand Alizadeh Alamdari 路 Toryn Klassen 路 Rodrigo Toro Icarte 路 Sheila McIlraith 馃敆 |
-
|
Robust Reinforcement Learning for Shifting Dynamics During Deployment
(
Poster
)
>
|
Samuel Stanton 路 Rasool Fakoor 路 Jonas Mueller 路 Andrew Gordon Wilson 路 Alexander Smola 馃敆 |
-
|
Uncertainty-based Safety-Critical Control using Bayesian Methods
(
Poster
)
>
|
Carlos Montenegro 路 Santiago Jimenez Leudo 路 Carlos Rodr铆guez 馃敆 |
-
|
Who Is the Strongest Enemy? Towards Optimal and Efficient Evasion Attacks in Deep RL
(
Poster
)
>
SlidesLive Video |
Yanchao Sun 路 Ruijie Zheng 路 Yongyuan Liang 路 Furong Huang 馃敆 |
-
|
What Would the Expert do()?: Causal Imitation Learning
(
Poster
)
>
SlidesLive Video |
Gokul Swamy 路 Sanjiban Choudhury 路 James Bagnell 路 Steven Wu 馃敆 |
-
|
Reinforcement Learning with Feedback from Multiple Humans with Diverse Skills
(
Poster
)
>
SlidesLive Video |
Taku Yamagata 路 Ryan McConville 路 Raul Santos-Rodriguez 馃敆 |
-
|
Efficiently Improving the Robustness of RL Agents against Strongest Adversaries
(
Poster
)
>
SlidesLive Video |
Yongyuan Liang 路 Yanchao Sun 路 Ruijie Zheng 路 Furong Huang 馃敆 |
-
|
Learning Contraction Policies from Offline Data
(
Poster
)
>
SlidesLive Video |
Navid Rezazadeh 路 Negar Mehr 馃敆 |
-
|
Safety-guaranteed trajectory planning and control based on GP estimation for unmanned surface vessels
(
Poster
)
>
SlidesLive Video |
Shuhao Zhang 路 Seth Siriya 馃敆 |