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 …

Deep neural networks (DNN) have been shown to be notoriously brittle to small
perturbations in their input data. This problem is analogous to the over-fitting problem in test …

The past decade has seen the great potential of applying deep neural network (DNN) based
software to safety-critical scenarios, such as autonomous driving. Similar to traditional …

Detecting bugs in deep learning software at the architecture level provides additional
benefits that detecting bugs at the model level does not provide. This paper makes the first …

Recent effort to test deep learning systems has produced an intuitive and compelling test
criterion called neuron coverage (NC), which resembles the notion of traditional code …

Input constraints are useful for many software development tasks. For example, input
constraints of a function enable the generation of valid inputs, ie, inputs that follow these …

Deep learning (DL) techniques are proven effective in many challenging tasks, and become
widely-adopted in practice. However, previous work has shown that DL libraries, the basis of …

Deep learning (DL) systems are increasingly applied to safety-critical domains such as
autonomous driving cars. It is of significant importance to ensure the reliability and …

Deep learning (DL) techniques attract people from various fields with superior performance
in making progressive breakthroughs. To ensure the quality of DL techniques, researchers …

Deep neural networks (DNNs) have been deployed in a wide range of applications. We
introduce a DNN testing and debugging tool, called DeepConcolic, which is able to detect …