Refactoring a DNN by dropping or scaling layers and distilling the original's outputs can make safety properties checkable, at the cost of verifying a surrogate rather than the original network.
Autonomous naviga- tion of uav by using real-time model-based reinforcement learning,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.NE 1years
2019 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Refactoring Neural Networks for Verification
Refactoring a DNN by dropping or scaling layers and distilling the original's outputs can make safety properties checkable, at the cost of verifying a surrogate rather than the original network.