On-Off Pattern Encoding and Path-Count Encoding as Deep Neural Network Representations (2401.09518v1)
Abstract: Understanding the encoded representation of Deep Neural Networks (DNNs) has been a fundamental yet challenging objective. In this work, we focus on two possible directions for analyzing representations of DNNs by studying simple image classification tasks. Specifically, we consider \textit{On-Off pattern} and \textit{PathCount} for investigating how information is stored in deep representations. On-off pattern of a neuron is decided as on' oroff' depending on whether the neuron's activation after ReLU is non-zero or zero. PathCount is the number of paths that transmit non-zero energy from the input to a neuron. We investigate how neurons in the network encodes information by replacing each layer's activation with On-Off pattern or PathCount and evaluating its effect on classification performance. We also examine correlation between representation and PathCount. Finally, we show a possible way to improve an existing DNN interpretation method, Class Activation Map (CAM), by directly utilizing On-Off or PathCount.
- Hervé Abdi. The kendall rank correlation coefficient. Encyclopedia of Measurement and Statistics. Sage, Thousand Oaks, CA, pages 508–510, 2007.
- Towards better understanding of gradient-based attribution methods for deep neural networks. arXiv preprint arXiv:1711.06104, 2017.
- Improving deep neural networks for lvcsr using rectified linear units and dropout. In 2013 IEEE international conference on acoustics, speech and signal processing, pages 8609–8613. IEEE, 2013.
- Visualizing higher-layer features of a deep network. University of Montreal, 1341(3):1, 2009.
- Complexity of linear regions in deep networks. arXiv preprint arXiv:1901.09021, 2019.
- Deep relu networks have surprisingly few activation patterns. In Advances in Neural Information Processing Systems, pages 361–370, 2019.
- Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 770–778, 2016.
- Mobilenets: Efficient convolutional neural networks for mobile vision applications. arXiv preprint arXiv:1704.04861, 2017.
- Imagenet classification with deep convolutional neural networks. In Advances in neural information processing systems, pages 1097–1105, 2012.
- Does resnet learn good general purpose features? In Proceedings of the 2017 International Conference on Artificial Intelligence, Automation and Control Technologies, pages 1–5, 2017.
- Inception and resnet features are (almost) equivalent. Cognitive Systems Research, 59:312–318, 2020.
- On the number of linear regions of deep neural networks. In Advances in neural information processing systems, pages 2924–2932, 2014.
- Rectified linear units improve restricted boltzmann machines. In ICML, 2010.
- Activation functions: Comparison of trends in practice and research for deep learning. arXiv preprint arXiv:1811.03378, 2018.
- On the number of response regions of deep feed forward networks with piece-wise linear activations. arXiv preprint arXiv:1312.6098, 2013.
- Adversarial defense through network profiling based path extraction. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 4777–4786, 2019.
- Searching for activation functions. arXiv preprint arXiv:1710.05941, 2017.
- Evaluating the visualization of what a deep neural network has learned. IEEE transactions on neural networks and learning systems, 28(11):2660–2673, 2016.
- Restricting the flow: Information bottlenecks for attribution. arXiv preprint arXiv:2001.00396, 2020.
- Grad-cam: Visual explanations from deep networks via gradient-based localization. In Proceedings of the IEEE international conference on computer vision, pages 618–626, 2017.
- Sagar Sharma. Activation functions in neural networks. Towards Data Science, 6, 2017.
- Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556, 2014.
- Going deeper with convolutions. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 1–9, 2015.
- Interpret neural networks by identifying critical data routing paths. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 8906–8914, 2018.
- Visualizing and comparing convolutional neural networks. arXiv preprint arXiv:1412.6631, 2014.
- Visualizing and understanding convolutional networks. In European conference on computer vision, pages 818–833. Springer, 2014.
- On rectified linear units for speech processing. In 2013 IEEE International Conference on Acoustics, Speech and Signal Processing, pages 3517–3521. IEEE, 2013.
- Learning deep features for discriminative localization. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 2921–2929, 2016.