Hide-and-Seek: A Template for Explainable AI

2020/04/30 by Thanos Tagaris, Tagaris, Thanos, Andreas Stafylopatis +1
Computer Science · Mathematics · #62M45 #Adversarial Robustness in Machine Learning #Anomaly Detection Techniques and Applications #Artificial Intelligence (cs.AI) #Explainable Artificial Intelligence (XAI) #FOS: Computer and information sciences #I.2.6 #Machine Learning (cs.LG) #Machine Learning (stat.ML) #acm:62M45 #cs.AI #cs.LG #msc:62M45 #stat.ML

paper · pdf · doi:10.48550/arxiv.2005.00130

24 pages, 14 figures. Submitted on a special issue for Explainable AI, on Elsevier's "Artificial Intelligence"

arxiv created 2020/04/30 · openalex publication_date 2020/04/30 · arxiv updated 2020/05/04 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Lack of transparency has been the Achilles heal of Neural Networks and their wider adoption in industry. Despite significant interest this shortcoming has not been adequately addressed. This study proposes a novel framework called Hide-and-Seek (HnS) for training Interpretable Neural Networks and establishes a theoretical foundation for exploring and comparing similar ideas. Extensive experimentation indicates that a high degree of interpretability can be imputed into Neural Networks, without sacrificing their predictive power.

Citations

• Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
• Distilling the Knowledge in a Neural Network
• Recurrent Models of Visual Attention
• Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computation
• Intelligent Agent Transparency in Human–Agent Teaming for Multi-UxV Management
• Hierarchical Multiscale Recurrent Neural Networks
• Axiomatic Attribution for Deep Networks
• A Theoretical Explanation for Perplexing Behaviors of Backpropagation-based Visualizations
• Sanity Checks for Saliency Maps
• Improving Semantic Segmentation via Video Propagation and Label Relaxation
• XLNet: Generalized Autoregressive Pretraining for Language Understanding
• Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps
• On a Method to Measure Supervised Multiclass Model’s Interpretability: Application to Degradation Diagnosis (Short Paper)

Related

• Explaining Deep Neural Networks by Leveraging Intrinsic Methods
• Toward Transparent AI: A Survey on Interpreting the Inner Structures of Deep Neural Networks
• Explaining Deep Neural Networks and Beyond: A Review of Methods and Applications
• Unboxing the Black Box: Mechanistic Interpretability for Algorithmic Understanding of Neural Networks
• How to Explain Neural Networks: an Approximation Perspective