Search papers, labs, and topics across Lattice.
TU Berlin, BIFOLD-Berlin Institute for the Foundations of Learning and Data
2
0
4
GDCE-I achieves faithful and interpretable counterfactual explanations for graph neural networks without compromising on the integrity of the data structure or the search space.
XAI-based techniques like Counterfactual Knowledge Distillation (CFKD) can significantly boost DNN generalization by mitigating spurious correlations, but their practical deployment is often hampered by the need for group labels and the unreliability of model selection in imbalanced datasets.
