Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox
Picture for Dylan Hadfield-Menell

Dylan Hadfield-Menell

Estimating and Penalizing Induced Preference Shifts in Recommender Systems


Apr 25, 2022
Micah Carroll, Dylan Hadfield-Menell, Stuart Russell, Anca Dragan

Figure 1 for Estimating and Penalizing Induced Preference Shifts in Recommender Systems
Figure 2 for Estimating and Penalizing Induced Preference Shifts in Recommender Systems
Figure 3 for Estimating and Penalizing Induced Preference Shifts in Recommender Systems
Figure 4 for Estimating and Penalizing Induced Preference Shifts in Recommender Systems

  Access Paper or Ask Questions

Linguistic communication as (inverse) reward design


Apr 11, 2022
Theodore R. Sumers, Robert D. Hawkins, Mark K. Ho, Thomas L. Griffiths, Dylan Hadfield-Menell

Figure 1 for Linguistic communication as (inverse) reward design
Figure 2 for Linguistic communication as (inverse) reward design
Figure 3 for Linguistic communication as (inverse) reward design
* 6 pages, 3 figures. Accepted at Learning from Natural Language Supervision workshop (ACL 2022) 
  Access Paper or Ask Questions

Guided Imitation of Task and Motion Planning


Dec 06, 2021
Michael James McDonald, Dylan Hadfield-Menell

Figure 1 for Guided Imitation of Task and Motion Planning
Figure 2 for Guided Imitation of Task and Motion Planning
Figure 3 for Guided Imitation of Task and Motion Planning
Figure 4 for Guided Imitation of Task and Motion Planning
* 16 pages, 6 figures, 2 tables, submitted to Conference on Robot Learning 2021, to be published in Proceedings of Machine Learning Research 
  Access Paper or Ask Questions

What are you optimizing for? Aligning Recommender Systems with Human Values


Jul 22, 2021
Jonathan Stray, Ivan Vendrov, Jeremy Nixon, Steven Adler, Dylan Hadfield-Menell

* Originally presented at the ICML 2020 Participatory Approaches to Machine Learning workshop 
  Access Paper or Ask Questions

Consequences of Misaligned AI


Feb 07, 2021
Simon Zhuang, Dylan Hadfield-Menell

Figure 1 for Consequences of Misaligned AI
Figure 2 for Consequences of Misaligned AI
* NeurIPS 2020 
  Access Paper or Ask Questions

Multi-Principal Assistance Games: Definition and Collegial Mechanisms


Dec 29, 2020
Arnaud Fickinger, Simon Zhuang, Andrew Critch, Dylan Hadfield-Menell, Stuart Russell

* arXiv admin note: text overlap with arXiv:2007.09540 
  Access Paper or Ask Questions

Multi-Principal Assistance Games


Jul 19, 2020
Arnaud Fickinger, Simon Zhuang, Dylan Hadfield-Menell, Stuart Russell

Figure 1 for Multi-Principal Assistance Games

  Access Paper or Ask Questions

Silly rules improve the capacity of agents to learn stable enforcement and compliance behaviors


Jan 25, 2020
Raphael Köster, Dylan Hadfield-Menell, Gillian K. Hadfield, Joel Z. Leibo

Figure 1 for Silly rules improve the capacity of agents to learn stable enforcement and compliance behaviors
Figure 2 for Silly rules improve the capacity of agents to learn stable enforcement and compliance behaviors
Figure 3 for Silly rules improve the capacity of agents to learn stable enforcement and compliance behaviors
Figure 4 for Silly rules improve the capacity of agents to learn stable enforcement and compliance behaviors

  Access Paper or Ask Questions

An Extensible Interactive Interface for Agent Design


Jun 10, 2019
Matthew Rahtz, James Fang, Anca D. Dragan, Dylan Hadfield-Menell

Figure 1 for An Extensible Interactive Interface for Agent Design
Figure 2 for An Extensible Interactive Interface for Agent Design
Figure 3 for An Extensible Interactive Interface for Agent Design
Figure 4 for An Extensible Interactive Interface for Agent Design
* Presented at 2019 ICML Workshop on Human in the Loop Learning (HILL 2019), Long Beach, USA 
  Access Paper or Ask Questions

Adversarial Training with Voronoi Constraints


May 02, 2019
Marc Khoury, Dylan Hadfield-Menell

Figure 1 for Adversarial Training with Voronoi Constraints
Figure 2 for Adversarial Training with Voronoi Constraints
Figure 3 for Adversarial Training with Voronoi Constraints
Figure 4 for Adversarial Training with Voronoi Constraints
* arXiv admin note: substantial text overlap with arXiv:1811.00525 
  Access Paper or Ask Questions

Conservative Agency via Attainable Utility Preservation


Feb 26, 2019
Alexander Matt Turner, Dylan Hadfield-Menell, Prasad Tadepalli

Figure 1 for Conservative Agency via Attainable Utility Preservation
Figure 2 for Conservative Agency via Attainable Utility Preservation
Figure 3 for Conservative Agency via Attainable Utility Preservation
Figure 4 for Conservative Agency via Attainable Utility Preservation

  Access Paper or Ask Questions

The Assistive Multi-Armed Bandit


Jan 24, 2019
Lawrence Chan, Dylan Hadfield-Menell, Siddhartha Srinivasa, Anca Dragan

Figure 1 for The Assistive Multi-Armed Bandit
Figure 2 for The Assistive Multi-Armed Bandit
Figure 3 for The Assistive Multi-Armed Bandit
Figure 4 for The Assistive Multi-Armed Bandit
* Accepted to HRI 2019 
  Access Paper or Ask Questions

On the Utility of Model Learning in HRI


Jan 04, 2019
Rohan Choudhury*, Gokul Swamy*, Dylan Hadfield-Menell, Anca Dragan

Figure 1 for On the Utility of Model Learning in HRI
Figure 2 for On the Utility of Model Learning in HRI
Figure 3 for On the Utility of Model Learning in HRI
Figure 4 for On the Utility of Model Learning in HRI

  Access Paper or Ask Questions

Human-AI Learning Performance in Multi-Armed Bandits


Dec 21, 2018
Ravi Pandya, Sandy H. Huang, Dylan Hadfield-Menell, Anca D. Dragan

Figure 1 for Human-AI Learning Performance in Multi-Armed Bandits
Figure 2 for Human-AI Learning Performance in Multi-Armed Bandits
Figure 3 for Human-AI Learning Performance in Multi-Armed Bandits
Figure 4 for Human-AI Learning Performance in Multi-Armed Bandits
* Artificial Intelligence, Ethics and Society (AIES) 2019 
  Access Paper or Ask Questions

Legible Normativity for AI Alignment: The Value of Silly Rules


Nov 03, 2018
Dylan Hadfield-Menell, McKane Andrus, Gillian K. Hadfield

Figure 1 for Legible Normativity for AI Alignment: The Value of Silly Rules
Figure 2 for Legible Normativity for AI Alignment: The Value of Silly Rules
Figure 3 for Legible Normativity for AI Alignment: The Value of Silly Rules

  Access Paper or Ask Questions

On the Geometry of Adversarial Examples


Nov 01, 2018
Marc Khoury, Dylan Hadfield-Menell

Figure 1 for On the Geometry of Adversarial Examples
Figure 2 for On the Geometry of Adversarial Examples
Figure 3 for On the Geometry of Adversarial Examples
Figure 4 for On the Geometry of Adversarial Examples

  Access Paper or Ask Questions

Active Inverse Reward Design


Sep 09, 2018
Sören Mindermann, Rohin Shah, Adam Gleave, Dylan Hadfield-Menell

Figure 1 for Active Inverse Reward Design
Figure 2 for Active Inverse Reward Design
Figure 3 for Active Inverse Reward Design
* ICML/IJCAI/AAMAS 2018 Workshop: Goals for Reinforcement Learning 
  Access Paper or Ask Questions

An Efficient, Generalized Bellman Update For Cooperative Inverse Reinforcement Learning


Jun 11, 2018
Dhruv Malik, Malayandi Palaniappan, Jaime F. Fisac, Dylan Hadfield-Menell, Stuart Russell, Anca D. Dragan

Figure 1 for An Efficient, Generalized Bellman Update For Cooperative Inverse Reinforcement Learning
Figure 2 for An Efficient, Generalized Bellman Update For Cooperative Inverse Reinforcement Learning
Figure 3 for An Efficient, Generalized Bellman Update For Cooperative Inverse Reinforcement Learning
Figure 4 for An Efficient, Generalized Bellman Update For Cooperative Inverse Reinforcement Learning

  Access Paper or Ask Questions

Simplifying Reward Design through Divide-and-Conquer


Jun 07, 2018
Ellis Ratner, Dylan Hadfield-Menell, Anca D. Dragan

Figure 1 for Simplifying Reward Design through Divide-and-Conquer
Figure 2 for Simplifying Reward Design through Divide-and-Conquer
Figure 3 for Simplifying Reward Design through Divide-and-Conquer
Figure 4 for Simplifying Reward Design through Divide-and-Conquer
* Robotics: Science and Systems (RSS) 2018 
  Access Paper or Ask Questions

Incomplete Contracting and AI Alignment


Apr 12, 2018
Dylan Hadfield-Menell, Gillian Hadfield

Figure 1 for Incomplete Contracting and AI Alignment
Figure 2 for Incomplete Contracting and AI Alignment

  Access Paper or Ask Questions

Pragmatic-Pedagogic Value Alignment


Feb 05, 2018
Jaime F. Fisac, Monica A. Gates, Jessica B. Hamrick, Chang Liu, Dylan Hadfield-Menell, Malayandi Palaniappan, Dhruv Malik, S. Shankar Sastry, Thomas L. Griffiths, Anca D. Dragan

Figure 1 for Pragmatic-Pedagogic Value Alignment
Figure 2 for Pragmatic-Pedagogic Value Alignment
* International Symposium on Robotics Research, 2017 
 * Published at the International Symposium on Robotics Research (ISRR 2017) 
  Access Paper or Ask Questions

Expressive Robot Motion Timing


Feb 05, 2018
Allan Zhou, Dylan Hadfield-Menell, Anusha Nagabandi, Anca D. Dragan

Figure 1 for Expressive Robot Motion Timing
Figure 2 for Expressive Robot Motion Timing
Figure 3 for Expressive Robot Motion Timing
Figure 4 for Expressive Robot Motion Timing
* HRI '17 Proceedings of the 2017 ACM/IEEE International Conference on Human-Robot Interaction Pages 22-31 
  Access Paper or Ask Questions

Inverse Reward Design


Nov 08, 2017
Dylan Hadfield-Menell, Smitha Milli, Pieter Abbeel, Stuart Russell, Anca Dragan

Figure 1 for Inverse Reward Design
Figure 2 for Inverse Reward Design
Figure 3 for Inverse Reward Design
Figure 4 for Inverse Reward Design
* Advances in Neural Information Processing Systems 30 (NIPS 2017) 
  Access Paper or Ask Questions

The Off-Switch Game


Jun 16, 2017
Dylan Hadfield-Menell, Anca Dragan, Pieter Abbeel, Stuart Russell

Figure 1 for The Off-Switch Game
Figure 2 for The Off-Switch Game
Figure 3 for The Off-Switch Game
Figure 4 for The Off-Switch Game

  Access Paper or Ask Questions

Should Robots be Obedient?


May 28, 2017
Smitha Milli, Dylan Hadfield-Menell, Anca Dragan, Stuart Russell

Figure 1 for Should Robots be Obedient?
Figure 2 for Should Robots be Obedient?
Figure 3 for Should Robots be Obedient?
Figure 4 for Should Robots be Obedient?
* Accepted to IJCAI 2017 
  Access Paper or Ask Questions

Cooperative Inverse Reinforcement Learning


Nov 12, 2016
Dylan Hadfield-Menell, Anca Dragan, Pieter Abbeel, Stuart Russell

Figure 1 for Cooperative Inverse Reinforcement Learning
Figure 2 for Cooperative Inverse Reinforcement Learning

  Access Paper or Ask Questions