Joar Skalse

According to our database, Joar Skalse authored at least 18 papers between 2019 and 2024.

Collaborative distances:

• Dijkstra number of four.
• Erdős number of three.

2024

The Perils of Optimizing Learned Reward Functions: Low Training Error Does Not Guarantee Low Regret.

Lukas Fluri, Leon Lang, Alessandro Abate, Patrick Forré, David Krueger, Joar Skalse

CoRR, 2024

Towards Guaranteed Safe AI: A Framework for Ensuring Robust and Reliable AI Systems.

David Dalrymple, Joar Skalse, Yoshua Bengio, Stuart Russell, Max Tegmark, Sanjit Seshia, Steve Omohundro, Christian Szegedy, Ben Goldhaber, Nora Ammann, Alessandro Abate, Joe Halpern, Clark W. Barrett, Ding Zhao, Tan Zhi-Xuan, Jeannette Wing, Joshua B. Tenenbaum

CoRR, 2024

On the Expressivity of Objective-Specification Formalisms in Reinforcement Learning.

Rohan Subramani, Marcus Williams, Max Heitmann, Halfdan Holm, Charlie Griffin, Joar Max Viktor Skalse

Proceedings of the Twelfth International Conference on Learning Representations, 2024

STARC: A General Framework For Quantifying Differences Between Reward Functions.

Joar Max Viktor Skalse, Lucy Farnik, Sumeet Ramesh Motwani, Erik Jenner, Adam Gleave, Alessandro Abate

Proceedings of the Twelfth International Conference on Learning Representations, 2024

Quantifying the Sensitivity of Inverse Reinforcement Learning to Misspecification.

Joar Max Viktor Skalse, Alessandro Abate

Proceedings of the Twelfth International Conference on Learning Representations, 2024

Goodhart's Law in Reinforcement Learning.

Jacek Karwowski, Oliver Hayman, Xingjian Bai, Klaus Kiendlhofer, Charlie Griffin, Joar Max Viktor Skalse

Proceedings of the Twelfth International Conference on Learning Representations, 2024

2023

On the limitations of Markovian rewards to express multi-objective, risk-sensitive, and modal tasks.

Joar Skalse, Alessandro Abate

Proceedings of the Uncertainty in Artificial Intelligence, 2023

Invariance in Policy Optimisation and Partial Identifiability in Reward Learning.

Joar Max Viktor Skalse, Matthew Farrugia-Roberts, Stuart Russell, Alessandro Abate, Adam Gleave

Proceedings of the International Conference on Machine Learning, 2023

Misspecification in Inverse Reinforcement Learning.

Joar Skalse, Alessandro Abate

Proceedings of the Thirty-Seventh AAAI Conference on Artificial Intelligence, 2023

2022

Defining and Characterizing Reward Hacking.

Joar Skalse, Nikolaus H. R. Howe, Dmitrii Krasheninnikov, David Krueger

CoRR, 2022

Defining and Characterizing Reward Gaming.

Joar Skalse, Nikolaus H. R. Howe, Dmitrii Krasheninnikov, David Krueger

Proceedings of the Advances in Neural Information Processing Systems 35: Annual Conference on Neural Information Processing Systems 2022, 2022

Lexicographic Multi-Objective Reinforcement Learning.

Joar Skalse, Lewis Hammond, Charlie Griffin, Alessandro Abate

Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence, 2022

2021

Is SGD a Bayesian sampler? Well, almost.

Chris Mingard, Guillermo Valle Pérez, Joar Skalse, Ard A. Louis

J. Mach. Learn. Res., 2021

A General Counterexample to Any Decision Theory and Some Responses.

Joar Skalse

CoRR, 2021

Reinforcement Learning in Newcomblike Environments.

James Bell, Linda Linsefors, Caspar Oesterheld, Joar Skalse

Proceedings of the Advances in Neural Information Processing Systems 34: Annual Conference on Neural Information Processing Systems 2021, 2021

Safety Properties of Inductive Logic Programming.

Gavin Leech, Nandi Schoots, Joar Skalse

Proceedings of the Workshop on Artificial Intelligence Safety 2021 (SafeAI 2021) co-located with the Thirty-Fifth AAAI Conference on Artificial Intelligence (AAAI 2021), 2021

2019

Neural networks are a priori biased towards Boolean functions with low entropy.

Chris Mingard, Joar Skalse, Guillermo Valle Pérez, David Martínez-Rubio, Vladimir Mikulik, Ard A. Louis

CoRR, 2019

Risks from Learned Optimization in Advanced Machine Learning Systems.

Evan Hubinger, Chris van Merwijk, Vladimir Mikulik, Joar Skalse, Scott Garrabrant

CoRR, 2019