Quantifying AI Model Trust as a Model Sureness Measure by Bidirectional Active Processing and Visual Knowledge Discovery | Synapse