Dead-ends and Secure Exploration in Reinforcement Learning.

Mehdi Fatemi, Shikhar Sharma, Harm van Seijen, Samira Ebrahimi Kahou

Browse the full ICML paper archive.