A systematic literature review on ai safety: Identifying trends, challenges and future directions

W Salhab, D Ameyed, F Jaafar, H Mcheick - IEEE Access, 2024 - ieeexplore.ieee.org
Artificial intelligence (AI) is revolutionizing many aspects of our lives, except it raises
fundamental safety and ethical issues. In this survey paper, we review the current state of …

Shielded Reinforcement Learning: A review of reactive methods for safe learning

H Odriozola-Olalde, M Zamalloa… - 2023 IEEE/SICE …, 2023 - ieeexplore.ieee.org
Reinforcement Learning (RL) algorithms are showing promising results in simulated
environments, but their replication in real physical applications, even more so in safety …

Safe reinforcement learning via probabilistic logic shields

WC Yang, G Marra, G Rens, L De Raedt - ar** for human-robot collaboration with transparent matrix overlays
J Brawer, D Ghose, K Candon, M Qin… - Proceedings of the …, 2023 - dl.acm.org
One important aspect of effective human--robot collaborations is the ability for robots to
adapt quickly to the needs of humans. While techniques like deep reinforcement learning …

A learner-verifier framework for neural network controllers and certificates of stochastic systems

K Chatterjee, TA Henzinger, M Lechner… - … Conference on Tools and …, 2023 - Springer
Reinforcement learning has received much attention for learning controllers of deterministic
systems. We consider a learner-verifier framework for stochastic control systems and survey …

Symbolic task inference in deep reinforcement learning

H Hasanbeig, NY Jeppu, A Abate, T Melham… - Journal of Artificial …, 2024 - jair.org
This paper proposes DeepSynth, a method for effective training of deep reinforcement
learning agents when the reward is sparse or non-Markovian, but at the same time progress …

Correct-by-construction runtime enforcement in AI–A survey

B Könighofer, R Bloem, R Ehlers, C Pek - … to Thomas A. Henzinger on the …, 2022 - Springer
Runtime enforcement refers to the theories, techniques, and tools for enforcing correct
behavior with respect to a formal specification of systems at runtime. In this paper, we are …