On Modifying a Neural Network's Perception
This addresses the interpretability challenge for AI researchers and practitioners, though it appears incremental as it builds on existing methods for concept manipulation.
The paper tackles the problem of interpreting neural networks as black boxes by proposing a method to modify their perception of human-defined concepts, enabling the generation of hypothetical scenarios for understanding and debugging, with empirical evaluation on synthetic and ImageNet datasets showing that manipulations are well interpreted by models.
Artificial neural networks have proven to be extremely useful models that have allowed for multiple recent breakthroughs in the field of Artificial Intelligence and many others. However, they are typically regarded as black boxes, given how difficult it is for humans to interpret how these models reach their results. In this work, we propose a method which allows one to modify what an artificial neural network is perceiving regarding specific human-defined concepts, enabling the generation of hypothetical scenarios that could help understand and even debug the neural network model. Through empirical evaluation, in a synthetic dataset and in the ImageNet dataset, we test the proposed method on different models, assessing whether the performed manipulations are well interpreted by the models, and analyzing how they react to them.