Beyond adversarial examples: sampling and repairing diverse failures with RADIUM
Our method predicts a diverse set of safety-critical failure scenarios for learning-based autonomous systems, then repairs the control policy to reduce the severity of those failures.
摘要
评审与讨论
The double-blind review process has been compromised; consequently, I will not be providing a summary or review of the paper.
优点
The double-blind review process has been compromised; consequently, I will not be providing a summary or review of the paper.
缺点
The double-blind review process has been compromised; consequently, I will not be providing a summary or review of the paper.
问题
The double-blind review process has been compromised; consequently, I will not be providing a summary or review of the paper.
伦理问题详情
The double-blind review process has been compromised; consequently, I will not be providing a summary or review of the paper.
Is there a way that we could get some clarification on how the double-blind process was compromised? Is there a way that you or the meta-reviewer could share these concerns with us privately in a way that does not risk compromising the other reviewers? If there is some mistake in our paper that accidentally compromised the double-blind process, we would like to know about it so we can fix the issue.