Neural networks are difficult to interpret and debug. We introduce testing techniques for
neural networks that can discover errors occurring only for rare inputs. Specifically, we …

Abstract Machine learning models are notoriously difficult to interpret and debug. This is
particularly true of neural networks. In this work, we introduce automated software testing …

Neural networks are difficult to interpret and debug. We introduce testing techniques for
neural networks that can discover errors occurring only for rare inputs. Specifically, we …

The underlying software implementation of these operations may contain many branching
statements but many of these are based on the size of the matrix and thus the architecture of …

Neural networks are difficult to interpret and debug. We introduce testing techniques for
neural networks that can discover errors occurring only for rare inputs. Specifically, we …

TensorFuzz: Debugging Neural Networks with Coverage Guided Fuzzing Page 1 TensorFuzz:
Debugging Neural Networks with Coverage Guided Fuzzing Augustus Odena, Catherine …

Abstract Machine learning models are notoriously difficult to interpret and debug. This is
particularly true of neural networks. In this work, we introduce automated software testing …

Abstract Machine learning models are notoriously difficult to interpret and debug. This is
particularly true of neural networks. In this work, we introduce automated software testing …

Machine learning models are notoriously difficult to interpret and debug. This is particularly
true of neural networks. In this work, we introduce automated software testing techniques for …